Nice write up! I have been using classic back-prop neural networks since the 1980s, and deep learning for the last 8 years. This tech feels like a rocket ship that is accelerating exponentially!
I am in my 70s and I don't work much anymore. That said, I find myself spending many hours in a typical day doing what I call "gentleman scientist" activities around Large Language Models.
I was walking this morning with a non-tech friend and I was trying to impart to him my feeling that all knowledge work, and a lot of mundane work is getting disrupted. Even though I have many tech-friends who are "Large Language Model doubters", my friend got it this morning. Emotionally and psychologically I think that some tech and other highly paid knowledge workers just can't accept the sea-change that we are living through.
For myself, I am spending a very large amount of time experimenting with the OpenAI APIs, LangChain, and Llama-Index - and I am enjoying myself tremendously.
> Emotionally and psychologically I think that some tech and other highly paid knowledge workers just can't accept the sea-change that we are living through.
This is a good way to frame it. As a mid-career person, I’m trying to figure out how to respond to these developments in my own career.
I think there’s a good chance that software development as a career looks very different in 10 years in terms of the day to day work and opportunities for employment.
25 years ago in university, I did roughly the same stuff as I do today. Languages are higher on the abstraction ladder, networks and hardware are faster, etc, but stuff is still roughly the same. I can see a real discontinuity in the work on the horizon.
I've been calling the mockery and dismissiveness, "nervous laughter". I think we have a sharp crowd here but they either lack imagination or are in denial.
I do not think that is quite true. Most of what I am seeing is hesitance and wariness.
Every innovation can be used both for good and for evil, and it often has unintended side-effects. Just ask Thomas Midgley Jr., who played a major role in both leaded gasoline and CFCs. Or Alfred Nobel, who tried to make mining safer, only to be appalled at an erroneous obituary to describe him as the "merchant of death" due to its military use.
Over the last two decades we've seen tech giants go from a bunch of nerds in garages to parasites invading our personal lives and consuming every bit of data about us they can get their hands on - and social media went from a cute gimmick to a society-destroying tool powering genocides.
AI is undeniably extremely promising, and it is not hard to see that it has the potential to change the world forever. However, we are inherently unable to understand how it works, or to alter it in a meaningful way. It's a Pandora's Black Box: will it give us a better and easier life, or will it destroy us? Nobody knows, least of all the people who are opening it!
We're at a point in time where almost any novel technology is just going to take jobs away, increase wealth inequality, and down the line make the world a worse place. It's difficult to be happy about a new powerful tool finding its way into the hands of tech giants.
Yes, this stuff is fun to work with and use as an aid to programming. But I think what's dawning on a lot of us is that it seems to be able to facilitate a large increase productivity. And given that large productivity increase will industry need as many software developers (and other knowledge workers)?
Your opinion on this probably depends on what your job currently looks like. Everywhere I've worked there's been significantly more stuff we wanted to do than we could actually do. Speeding that up looks awesome to me.
But if you work at BigCo and mostly sit on your hands already, then, yeah. I don't know.
>Everywhere I've worked there's been significantly more stuff we wanted to do than we could actually do. Speeding that up looks awesome to me.
Does it? Slowing our breakneck anti-human business pace and stopping mindless and environmentally and culturally harmful consumption and production of trinkets is surely better!
Huh? Have you attended a local city council meeting recently? I can guarantee you they're running on shoestring budgets with an endless backlog of requests that they are only treading water on because of anemic tax revenue and low prestige for government jobs. "Productivity" isn't just some high tech, consumer bauble producing phenomenon. It helps everyone out. Highly profitable, headcount-bloated big companies are the outlier.
I can guarantee you that the reason they do so, is an explosion of BS requests, extreme bureucracy, and needless requirements, with a simultaneous drop in actual useful basic work - on city instrastructure and such.
Every BigCo I’ve worked at (and I’ve worked at places that put the B in Big) people aren’t sitting on their hands because there isn’t way more to do than they can feasibly do. It’s usually because they’re fundamentally illsuited for the job and sitting on their hands is better than being homeless, or the people who are illsuited for their jobs feel anxious about being homeless so force everyone into meetings and team building and off sites and retrospectives and sprint planning and burn down chart analysis and whatever else they can conceive of that seems productive but doesn’t change production in any way. To that extent at least an AI would be left alone to write actual code, and those folks can have their endless meetings and still justify their job the same way they ever did. No AI will ever host a breakout session at an offsite to discuss strategy for getting people to work in the horrific hotel seating dehumanized office, they’re too capable.
> No AI will ever host a breakout session at an offsite to discuss strategy for getting people to work in the horrific hotel seating dehumanized office
I can already feel it dramatically increasing my personal productivity. I expect it's going to eventually lead to a 2x or 3x time productivity increase for software engineering generally (if not more).
Will that result in less demand for software engineers? I doubt it. I think it will lead to companies and organizations doing way more software engineering.
Increased supply doesn't change the demand curve but there are basically no scenarios that don't lead to an increase in volume and a decrease in price.
Somewhat similar boat; I just hope it winds up helping me do my job better, instead of me trying to teach it how to do my job. Setting aside professional anxieties, after using chatGPT for a bit I quickly realized I am completely uninterested in prompt engineering. But as a tool, like with an IDE, it could be a big leap forward.
I literally just finished writing some documentation and had Copilot running, sometimes it'd autocomplete something with garbage (and I would just type what I wanted anyway), more than once it autocompleted with several lines of exactly what I wanted to write. This is way better than the phone-type autocomplete I've seen in gmail where it just guesses the next word.
I have started using it for a few days. Basically asking to write and modify pieces of code, for now small ones where I can easily spot a bullshit. I am very much looking forward for it to improve to the point where I can spend more time thinking about business domain rather then how to translate it to efficient code.
Indeed, it's fascinating to witness a sizable segment of the HN community and distinguished intellectuals like Chomsky displaying absolute denial. I've started bookmarking the comments so I can look back at them in 5 years and have a good laugh. Some from a few months back are already aging badly[0].
When Chomsky proposed generative grammars, his theory of universal language acquisition, and so on, they were radical ideas that upturned the central cannon of linguistics.
Time has been on his side - entire schools and subfields of linguistics went extinct as more evidence emerged that Chomsky was fundamentally right. Basically every computer language and data format in existence is parsed/lexed in ways inspired by his models of language.
But now Chomsky is considered the stodgy old establishment, and whenever one of his theories is contradicted somewhere on the margins people shout "Aha! He was wrong the whole time!" and ignore the 99% of cases where his models are still the best ones we have.
This is a hot take, would love a link/explanation on why you think neural nets are human-like. For example, from the op Ed you’re likely referencing:
“ Their deepest flaw is the absence of the most critical capacity of any intelligence: to say not only what is the case, what was the case and what will be the case — that’s description and prediction — but also what is not the case and what could and could not be the case. Those are the ingredients of explanation, the mark of true intelligence.”
Isn’t that just plainly true of LLMs? Sure they can produce text that looks like an explanation, and sure putting in a header telling it to spell out its reasoning will get it to output text that looks like reasoning, but I feel the nature of its hallucinations make it clear that it’s not actually performing those steps anywhere in the net.
But maybe I’m just a Luddite? These technologies are amazing and transformative, I’m just shocked to see hate on HN of CHOMSKY of all people, the father of modern linguistics and cognitive science…
Why do you think this is impossible? Have you tried making an account on chat.openai.com? Because it can in fact do that right now; and even people who are otherwise skeptical of LLMs have panned Chomsky's article.
He wrote an op ed recently with the oft-repeated argument that stochastic models are fundamentally distinct from structured symbolic/logical models. In simple terms you can never really “trust” chatgpt’s answers because it’s just guessing the answer that looks right, not applying structured reasoning like humans do.
> In simple terms you can never really “trust” chatgpt’s answers because it’s just guessing the answer that looks right, not applying structured reasoning like humans do.
That example was highly controversial, and I think Chomsky was actually right in that instance (all languages have recursion). Generally my understanding is that he's considered to have held the field of linguistics back quite a bit with a non-disprovable hypothesis.
He supported the communist though. He also supported Pol Pot. You can predict his opinions of foreign policy if you understand that he bases them on the "US is evil" axiom and therefore anyone who opposes it is good, no matter how actually evil they may be.
Chomsky absolutely was a Cambodian Genocide denailist at the time. He very arrogantly dismissed actual eyewitness testimony of the genocide from refugees purely for ideological reasons.
Did you skip my first link or are you disagreeing with it? It's about the same subject. And in your WP link has a second quote writing about the genocide as fact.
Chomsky's skepticism in the media criticism article talking (among other things) about a book about the then-contemporary events of course looks quite unfortunate in retrospect, but given the volume of work and constant writing about world events over a long career it seems to be a error in judgement in the moment, rather than denying history, especially as he evidently updated his views later as more evidence came about.
It was completely unjustified at the time. There was absolutely no reason to doubt the testimony of refugees coming from Cambodia, other than they didn't fit into Chomsky's strange worldview that the US is the cause of all problems in the world.
Even today his views on Russia's invasion of Ukraine are completely moronic.
For ‘Peace Activists,’ War Is About America, Never Russia
Their own hard-left worldview is so absorbing that they will take the side of any aggressor in the anti-Western camp.
I didn't bookmark them, but I recall comments from ten years ago predicting that 50% of the workforce was going to be replaced by AI in the next ten years. Fast-forward to "the future" and the reality is AI has left without a job a grand total of 0 people. The grandiose promises we're hearing now concerning AI are nothing new. They were laughable then, as they're now.
Those comments about employment are indeed laughable and I bookmark them as well. We're at near full employment despite the invention of mechanized farming, electricity, the printing press, cars, the computer, the Internet, etc. I was referring to people who are in denial about the rapid progress of AI, and the impact it is going to have in coming years.
[0] Until we reach AGI, but I don't dare attempt to predict when this is going to happen.
> Emotionally and psychologically I think that some tech and other highly paid knowledge workers just can't accept the sea-change that we are living through.
Soon to be 60 year old here. Glad that if need to I can retire now. Certainly the rise of LLMs and generative AI isn't going to be all bad, but I've also got a feeling that not as many software developers will be needed soon since those who can leverage LLMs will experience a pretty decent productivity boost. Part of me wonders if at least some of the layoffs we've seen in the last couple of months are because companies such as Google, Microsoft, Amazon, etc. (the ones that have pretty clear visibility into what's happening in the AI space) are realizing that they aren't going to need as many knowledge workers in the not-so-distant future.
I think there was always this idea in the back of our minds that this was going to happen someday. But someday was always like 15 to 20 years out. Looks like someday is knocking on our door now.
> Certainly the rise of LLMs and generative AI isn't going to be all bad, but I've also got a feeling that not as many software developers will be needed soon since those who can leverage LLMs will experience a pretty decent productivity boost.
OTOH, software development will become significantly cheaper. That means the business case for using software in more places just tilted in favor of throwing some software at it. You'll see more and more businesses using it, ones you didn't expect, like local mom and pop businesses.
Yes, it's going to be different but I don't think we know exactly what is going to happen yet.
Yeah definitely, but we already saw this coming before LLMs right? Low code solutions, site builders, I mean MS Frontpage is as old as the Web itself. Just like you needed top MechEs to build heat exchangers in the old days, nowadays a small team makes a design and technicians handle most of the actual work of assembly and maintenance. Likewise you'll probably just have low code engineers who glue existing parts together to make things work at your local cafe or restaurant.
One of the reasons the work at Big Tech is so fun is because they actually need to engineer a lot of things. Smaller tech companies or businesses that utilize tech can just glue a few libraries together to make it work.
What do you call the person they are going to hire to use the tool or are they going to switch gears from selling painted dodads and chicken sandwiches to fire up an IDE even a web based one.
The average human is only slightly smarter than the average rabbit. They are going to pay someone who specializes in this task.
Who needs people? Elites or not, what use will they bring? To do what? And elites do need plebs without which they’d have no point of reference to set themselves apart
One question out of academic curiosity: I'm exploring ways to use these tools for research projects in econ and am struggling to see a good angle.
For instance, suppose I have lots of PDF reports on how firms have evolved on each quarter (10,000 reports or any other number beyond what I can read).
Can LLMs be used to spit out variables based on these reports? EG: indicator variables (optimistic-vs-pesimistic), categories, etc. that I can then use as data to test economic models?
I tried simple cases with ChatGPT3.5 and got the feeling it's great for outputting narrative text, but felt mediocre for when the output was more narrowly defined into categories (which was surprising).
Look at the LangChain and Llama-Index (used to be called GPT-Index) projects that make smaller projects that need to use a large amount of text data do-able. There is also support for reading PDF files (and many other data sources), and pre-computing embeddings. If you spend a short while looking at example code in the documentation, find something that is similar to your requirements (e.g., semantic search, conversational chat about a set of documents, etc.), and build on that.
This caught my attention as I found it implausible:
> One DeepMind engineer even reported being able to convince ChatGPT that it was a Linux terminal and getting it to run some simple mathematical code to compute the first 10 prime numbers. Remarkably, it could finish the task faster than the same code running on a real Linux machine.
Following the link, there's a screenshot to a screenshot [0] of a code-golf solution to finding primes which is quite inefficient, and the author notes
> I want to note here that this codegolf python implementation to find prime numbers is very inefficient. It takes 30 seconds to evaluate the command on my machine, but it only takes about 10 seconds to run the same command on ChatGPT. So, for some applications, this virtual machine is already faster than my laptop.
So it's not quite calculating primes; more likely it recognizes the code as being code to do so, and recites the numbers from memory. That's interesting in its own right, but we won't be running Python on an LLM for a performance boost any time soon. In my experience this interpreting is apparent as a limitation of the model when it keeps insisting on broken code being correct, or having its mistake pointed out, then apologizing, saying it's got some new code that fixes the issue, and proceeding to output the exact same code.
So on the one hand, these newly publicized models can render convincing representations of realities we used to get from deterministic processes. On the other hand, it's probabilistic and fails to conform to logic quite often, and in a confident way.
We're building systems capable of programing computers non-deterministically. I think this is huge. But not because ChatGPT23 will be a CEO or a politician. But because this is a paradigm shift in compute similar to moving from integrator machines to general computers. I don't think LLMs will make programmers obsolete. I think large enough models will make programming something completely different from what it is now.
The days of sequencing tokens for compilers/interpreters seem to be drawing to an end as the dominant way of specifying software products.
The LLM can act as a global cache for common solutions to common problems, with the ability to perform the integration work necessary to apply them.
That prime number example is a little bit like when you put a functools.lru_cache decorator on a function in Python. It's faster than computing the function call because it's able to recall the return value for the parameters from the cache "memory".
Of course, many skilled programmers are also mainly used a cache for common solutions to common problems organizations have in the programming domain. As humans we can derive satisfaction from being able to tackle the same task others can, as a confirmation of our own progress. We like "doing it ourselves". But globally that's not super interesting if lots of people are constantly re-developing the same solutions to the same problems for their own benefit. I guess that's the push-and-pull (and the anxiety generator).
>The LLM can act as a global cache for common solutions to common problems, with the ability to perform the integration work necessary to apply them.
In my opinion Stack Overflow does a fine job at that. And it's transparent in terms of solutions proposed being voted on and discussed. Turning that into sausage advice is a downgrade.
>But globally that's not super interesting if lots of people are constantly re-developing the same solutions to the same problems for their own benefit.
I'd argue this is how we train individuals thus globally quite relevant.
I get the feeling. But I've always enjoyed the abstract puzzles more then anything. Computers attracted me as a form of very complex abstract puzzles.
But when it comes down to it everything in life is just nth dimensional tensor puzzles. What I really cherish computers for giving me is fast and clear feedback.
I'm about 10-15 years from retiring, and lately, I've been thinking a lot about how to handle the rest of my career with all these new AI advancements.
This disruption does appear to be different than others prior. As it is not a narrow disruption with limited boundaries for which we can plan and organize our lives for some stabilization period to follow.
Instead, it is constant and accelerating destabilization. Hundreds of AI projects attempting to ride this accelerating wave were essentially just made obsolete yesterday - https://www.youtube.com/watch?v=DH-2BHDYNfk
I feel the excitement is going to very soon turn into frustration of attempting to remain relevant ahead of the accelerating technological curve. Humans need periods of stabilization to plan and reason about their lives.
When I was a bored and under-challenged student in the early days of university, one of my tests during a programming exam was to write a program that spit out the factorial of an input integer.
For shits and giggles, I just just wrote a series of `if...then` statements for 0 to 5, and only actually calculated the factorial for inputs >=6. I passed after the examiner just input 3 and 5 and was satisfied that the output was correct. Oops.
Not 100% sure but I believe this is how we landed the lunar module on the moon the first time...tan/arctan/both were too hard to compute on the processors of those days so they discretized into half angles & stored the tangents in a lookup table.
I think its knowledge of python can only come from comments about code, since it can't execute the python or learn from trying code like a human would.
That doesn't change the fact that its predicting the output based on its knowledge of python though.
So, yes you're right, but at the same time it doesn't change anything IMO.
Large language models are storytellers. To write a story about someone using a computer terminal, there are things it’s helpful to know, but there are easier ways than simulating a computer.
Since we don’t know how it works, we should be open to the possibility that it’s using all sorts of storytelling techniques.
I did a few more experiments including generating large factorial numbers that took a long time on my laptop but it responded accurately to a much larger length than my laptop could do (though these were only accurate to the first few hundred digits)
Lol why does ChatGPT hit ^C? Do you think it’s getting bored of waiting to respond to you and decides that’s enough time spent on an answer?
Edit: the base64 decoding examples are terrifying. I have no idea how that works.
Edit2: actually I can sort of see how that works, b64 encoding doesn’t have any obfuscation to it, I could see how an ML model can build a pretty good approximation over time after seeing all the examples on the internet.
I was personally quite appalled that so many people believed this was what it was doing, but I suppose since the source was a DeepMind researcher, it lends a lot of credibility.
By the same reasoning, I hope people do realise that there are computations that are just impossible given the finite depth in a Transformer. The only possible way to overcome this in the current paradigm is the Chain-of-Thought related methods.
Yeah, I tried asking GPT-4 to write a simple Python program to find the millionth prime. It wrote the naive approach where you loop over numbers and for each number, check all of the possible divisors up to its square root, and then keep track of how many primes you've found. Obviously, this takes quite a while to execute for N = a million. And yet, if you ask GPT-4 what the output of the code will be, as if my magic (memorization) it immediately spits out 15485863. It makes it a lot more clear that it's not actually simulating the code.
yep I tried asking chatgpt to optimize some SQL queries with a heavy amount of full outer joins. The optimization I was trying to push it to was specifically adding an index on a column or filtering the first table that we join on but it kept creating SQL subqueries and switching the full joins to left joins no matter what I told it and the sql errors I sent it
Its speed as compared to the laptop - I assume it was running online, on OpenAI’s hardware. That hardware is quite extensive, I thought. I believe I’ve seen the cost of one (not ‘search’, what’s the word here, ‘output’?) is something like 1-10 cents.
I don't think "memoisation" is an accurate word for this; it implies doing the computation once, and storing the result in cache to return later. It's more like replacing your Python code with a SQL query of the LLM's understanding of what it's supposed to do, which may or may not be correct, and executing it on its "database" of knowledge, which may or may not have the correct data.
this is new to me but.. a quick read of Wikipedia [1] later, it appears that this decades-old method takes a goal, and then uses branching and recording to find paths that differ the least from the goal. The article mentions that the curse of dimensionality is so bad here that approximations are often used in practice. Does that capture it?
Equally interesting is the psychology of people who take the time to write long posts borderline gloating about tools making software engineers obsolete.
Okay, but you have another post in this thread with:
> When these LLMs get normalized probably 5 years from now I'm going go back to these old threads and contact these people who are in self denial and throw it in their face. I'll just link this comment and be like I TOLD YOU, I TOLD YOU, YOU WERE WRONG.
Not if it's said in a nice way. The commenter called LLMs a freaking auto complete. That's not nice either, but nice enough for HN. So I figured a little subtle gloating is deserved and well within the bounds of etiquette here on HN.
I'd like to see posts on LLMs written from a different perspective. For me, the surprise comes not from the sudden emergent capability of language models, but that the understanding (and synthesis!) of ideas encoded in language has succumbed to literally nothing more than statistical analysis. Or at least come that much closer to doing so.
That it bears so close a resemblance to actual thinking says more about the importance of language to cognition than the other way around.
This is what Stephen Wolfram concludes in a recent article about ChatGPT:
> The specific engineering of ChatGPT has made it quite compelling. But ultimately (at least until it can use outside tools) ChatGPT is “merely” pulling out some “coherent thread of text” from the “statistics of conventional wisdom” that it’s accumulated. But it’s amazing how human-like the results are. And as I’ve discussed, this suggests something that’s at least scientifically very important: that human language (and the patterns of thinking behind it) are somehow simpler and more “law like” in their structure than we thought.
This is key. ChatGPT/GPT-4 alone are limited to reformulating what they know from their training data. Linked to search engines, databases, and computational tools such as Wolfram Alpha, they acquire much more capability. We're already seeing that with Microsoft Bing.
(Update: what happens as large language models learn Excel?
Especially since Microsoft is already connecting them to Excel.)
What's striking is how fast this field is advancing. Huge advances over months, not years or decades.
We now have a much better idea of how intelligence evolved. It's mostly just more neurons. One of the great philosophical questions has, inadvertently, been answered.
> Linked to search engines, databases, and computational tools such as Wolfram Alpha, they acquire much more capability.
I see this as analogous to the human brain; there are different structures which are particularly good at specific tasks/functions. They all work together.
The only difference between a human brain and an ANN is a difference of degree. A neuron and an artificial neuron are functionally identical. I think as we start interconnecting these models we see surprising emergent properties.
I wonder if different languages lead to different capabilities? If I ask the same question in English, Japanese, and German, will I reliably get “better” answers from one language over another.
I find the phrase "statistical analysis" a frustrating one nowadays as it seems to have become a signal for "I hold a particular philosophy of the mind".
I don't understand this use of "statistical" as a diminutive to describe these models.
Why can't incredibly complicated behavior be emergent from matrix multiplication subject to optimization in the same way that our biological matter has developed complicated emergent properties also being subject to optimization?
The loss function is very different, the optimization techniques as well, but the fundamental idea of complex behavior emerging out of a substrate subject to optimization seems common. I haven't seen a single good answer to that
Well, to a degree, because we just don't do science like that.
You're the one supposing a thing, so the burden of proof is on you. You need to demonstrate that the "incredibly complicated behaviour" that you're referring to (I assume this is longhand for "thinking", but please correct me if I'm wrong) is indeed emerging from matrix multiplication. Especially given that what you're suggesting is unexpected, given the known way these models work and the explanations that have been put forth already that extrapolate from the known way these models work.
If science were so credulous as to accept the first proffered theory about a new development, well, we wouldn't have these interesting AI models in the first place!
> I find the phrase "statistical analysis" a frustrating one nowadays as it seems to have become a signal for "I hold a particular philosophy of the mind".
LLMs are trained to reproduce human text, that is different from for example AlphaGo that is trained to win Go games. Trained to reproduce data is what we mean with a statistical model, trained to win is how we got superhuman performance before, while trained to reproduce data performs worse than the original creators of the data.
This is reductive in the sense is that things like AlphaGo were bad at Go for a very long time, and then with more compute power and algorithm changes suddenly they were far better. And the problem space for go is absolutely huge it is still an insignificant portion of the problem space for knowledge.
Perhaps that's because you haven't asked a single good question?
What do matrix multiplication and optimisation have to do with the way the human mind, or the human brain work? That they have anything to do at all, is your assumption, that you seem absolutely convinced about- and then you go asking people why can't it be true? You say why it is true. It's your assumption.
Matrix multiplication and optimization are human mathematical techniques. They have about as good a chance of being something that exists in nature independently of humans as Magic: the Gathering and Call of Duty. They might be useful models to help us understand how things work, but to assume they are how things work is a huge leap of faith.
Ask the right questions and then an answer may even suggest itself. Ask questions that follow from your preconceived answers and you're in a world of fantasy.
Some anthropologists suggest that our main evolutionary advantage was not so much our individual ability for reasoning, but our collective capacity to accumulate, evolve and transmit cultural knowledge over the centuries and millennia.
Skills like fire or language, for example, had a major influence in the development of our species and are mainly culturally transmitted: trying to reason your way into creating one or the other from scratch is a surprisingly difficult task.
If that point of view is true, then it shouldn’t be surprising that a large part of what we consider human-like behaviours should be tractable simply by analysing large amounts of data. AI systems are not modelling cognition, but culture.
To use my favorite Stephen Hawking quotation (that's also been sampled by Pink Floyd), "for millions of years, mankind lived just like the animals, then something happened that unleashed the power of our imagination: we learned to talk."
I think the task of predicting the next word can be misunderstood. The better you want to be the more you have to "understand" how the previous words interacted. From the style of writing to the current topic discussed, the task gets increasingly complex if you want to be really, really good. How could the next sentence start? Will the author end the sentence here or keep going? These questions are very complex.
This does not mean that we humans might predict all the time, in fact I would argue that LLMs only predict during training. They generate otherwise. We might also learn by trying to predict. I can imagine babies doing it.
I had the same thought watching my son as a baby. So much of his day seemed to be focused on predicting what happens next, and he got so much joy when he succeeded. So many life skills are predicated on knowing that if I do this, that will happen, which gets me closer to my goal. I started to wonder if intelligence and prediction are really the same thing.
Interesting take.. especially as I often wondered, watching and loving animals, if not our ""intelligence separation"" from them is not to a large degree because of the language.. (and the wisdom storage and transfer that comes with it).
There is a startling acceleration of innovation in the field that GPT-4 illustrates. According to NVidia, LLM sizes have been increasing 10X per year for the last few years. This tech is going to hit every aspect of society like a sledgehammer over the next 48 months.
The same was said 10 years ago. It's astonishing what can be done, but you can already see fundamental limits. I think it will raise productivity for some tasks, but not fundamentally change society.
The number of people saying it now is many orders of magnitude more than the number of people saying it 10 years ago. Not saying that means it will happen, but it isn't the same situation.
The A.I. Monte Carlo Ray Tracing Hallucination Engine can change society by showing as-is and to-be next state. Two use-cases: new infrastructure installation or upgrade and time interval inspection tracking ideal-case vs real world condition. Suppose a lazy contractor skips steps and cuts corners, or a pathological contractor builds the thing and pulls the thing apart over and over again when all that was needed was a one and done, or the change is for the worse. A civil engineer can walk around with an iPad and iPhone to check-in the master plan.
You need to replace the civil engineer there (in today's world this is an inspector, I don't know for sure if they are professional engineers or not) for it to be useful, but you still need to have someone not working for the contractor who's incentivized to fake it.
The trouble with many of those construction examples is that they're point in time. Installer fucks up how the windows are installed re: flashing and water-proofing, but then puts the siding over it... the error is now completely hidden.
You could automated the inspection by AI photo analysis of every single window install on the project, say - but we could already do that for the inspection vs sending someone out, and send the photos to the expert instead, and we don't. Whether that's for lack of incentive to improve, or for ability to go deeper out of distrust for builders? I don't know.
Given a few sensors that could be installed on the house's pipes, one can imagine several ways an AI can check that all is good. Same for checking blueprints for quality/validity and purchases the plumber says are "needed"...
They may also run out of data if they already consumed most of the internet. Or start producing so much of the internet's content that LLMs start consuming what they write in a closed loop.
I would not be so sure about compute capacity? Neural network architectures are still in their infancy, it is very likely that more efficient approaches exist.
Sure, but the small & efficient networks so far have been pretty poor. Edge AI was hugely hyped a few years ago and it petered out. All of the big tech companies seems to be chasing the biggest networks with the largest training sets, that seems to be the direction for now.
There could be huge efficiency savings within the implementations but when training costs are already this high it seems naive to think the low-hanging fruit is still there.
I wouldn't be so confident. Flash attention is a recent, significant improvement to training times and sure looks low-hanging now.
I was not familiar with Edge AI, interesting concept. I feel like improving the efficiency of very large models is much more likely.
The recent successes will lead to an even larger influx of $ in the short term --- this is an existential threat to Google, after all. We will see where things go!
We've had neural nets since 1943. The architectures are not "in their infancy", new architectures have been developing for decades, even entire neural net paradigms (feed-forward nets, recurrent nets, recursive nets, etc etc.). Their scale has also been increasing ever since Hinton and friends rediscovered backprop in the '80s. Neural nets are positively ancient at this point, not "in their infancy"!
I don't know why people just keep repeating this complete fantasy as if it were true. Where does it originate from, I wonder? I suspect someone said something like that on social media, their post went viral, and now all of the internet is reverberating with this thing. It's a meme, yes?
> Where does it originate from, I wonder? I suspect someone said something like that on social media, their post went viral, and now all of the internet is reverberating with this thing. It's a meme, yes?
I don't have social media outside of HN.
It comes from a few observations:
1) Large models are still improving with increased parameter counts (we do not know where the ceiling is yet; it could be low but it could also be high).
2) Most current architectures train by using all model parameters to produce an output, which is vastly inefficient. While it is not clear how to improve on this in the general case yet, in the simpler problem of NERFs, sidestepping this issue has led to a ~100x improvement in training time.
3) https://mingukkang.github.io/GigaGAN/ very recently increased the parameter count of StyleGAN by selecting parameters dynamically at runtime. They improved on previous results by a very, very large margin, at somewhat comparable training times.
I stand by my claim: "neural network architectures are still in their infancy, it is very likely that more efficient approaches exist". I am not claiming that AI will become sentient or anything crazy and do not understand why you are associating my point of view with other people. I just said that it is likely that a novel technology will continue to improve (has this it ever NOT been the case for any new technology?).
Well, if you want to know whether neural nets are in their "infancy" you shouldn't make "observations", you should read the literature. It goes back many years. Go to the primary sources, why try to guess and risk guessing wrong, as here?
>> I stand by my claim: "neural network architectures are still in their infancy, it is very likely that more efficient approaches exist".
Half of your "claim" is incorrect. Don't just double down on it! There's really nothing to "claim" here, the "infancy" or not of neural nets is not a matter of claiming or guessing. Either you know what it is, or you don't.
Your tone comes off as very adversarial/angry. Intentional or not? The claim is about the likelihood of more efficient approaches existing, not about the subjective qualifier preceding it (misunderstanding?). Yes, NNs have existed for a while now, but they will also exist for a long time forward in the future (hence the perhaps poor choice of word: infancy).
"if you want to know whether neural nets are in their 'infancy' you shouldn't make 'observations'" -> I think you have understood this as "neural networks did not exist before" when what I meant is "neural networks will still change a lot in the future"
I was interested in discussing, specifically: will NNs continue scaling up in size in the near/long term. In answer to the statement "the training costs will exceed all the compute capacity in existence", I added "it is very likely that more efficient approaches exist.". You replied that NNs have existed for a long time, and that because of this, it is a complete fantasy that more efficient approaches exist. Do you feel like this is a fair assessment or no?
My problem is that I read what people write on HN and treat it with the same seriousness and respect I want people to treat my comments, when the majority are only saying whatever comes to their mind just to make some sort of impression. Then when I point out some obvious error, people freak out and get defensive because they never expected anyone to take them seriously, they're just spouting off whatever without really thinking. And then they try to wiggle out of the conversation, just like you're doing right now, pretending that you were misunderstood.
Well, my mistake then for taking you seriously. Many apologies. You can rest assured it won't happen again.
I feel very sad reading this. You are absolutely correct that I view HN as watercooler where I make hyperbolic statements, but incorrect when you say that that I did not take you seriously. I was frustrated because I felt like you were latching on to a figure of speech just to be argumentative instead of trying to understand what I meant (I was not trying to say that neural networks are literally new, I was trying to say that we have no proof that the parameter to flop ratio is even close to optimal because architectures are still changing a lot).
You are right that neural networks are not in their infancy. Best wishes.
LLMs of this type will just start making up shit when they dont know something. because they simply generate the most probable next token based on previous x tokens. this is not fixable.
this alone makes these LLMs practically unusable in vast majority of real-world applications where you would otherwise imagine this tech to be used.
for instance, your comment confidently states this is unfixable - presumably based on the frequency you've seen similar text on the internet. why should anyone believe the veracity of your statement? These things didn't have any of these emergent capabilities one year ago, why are you so sure you understand their nature one year from now?
"your comment confidently states this is unfixable - presumably based on the frequency you've seen similar text on the internet. why should anyone believe the veracity of your statement? "
Yep, the idea of grounding seems interesting to me. Everything in a LLM is just a statistical dream at this point with no 'reality basis' at this point. I wonder if it's possible to give the language model grounding points of things that are real and building a truth model from that.
The reason they seem to make things up is because they have no way to verify anything, they can only speak of things in relation to other things, but they have no epistemic framework. This is very much a fixable problem that augmentation with logic engines and a way to prioritise truth-claims could go some ways towards solving.
My memory could be improved by connecting my brain to an external hard drive. Wiring them together, alas, is not just hard; we have absolutely no idea how.
We do have some idea how, most people just don't really want to deal with the nightmare of being augmented and the life changing consequences that come with it, on top of the risk.
To me "we have no idea how" != "this is not fixable" (and even "we have no idea how" even seems like a strong statement.)
Perhaps it's because I'm ignorant about the inner workings, but calling the problem "unfixable" so early in the evolution of LLMs seems like foolish certainty.
I pretty much agree. That wasn't me who called it unfixable. (This is a UI issue with HN that keeps coming up for me -- people being mistaken for the OP.)
this is absolutely not a fundamental limit but simply a hard challenge. Approaches exist and it is an active field of research where we do make progress.
fair. It's not proven that a solution exists for our models but I don't see much that leads me to believe it's impossible. I know GPT is not reliable but there's also really not much done to improve reliability. Its open research but certainly interesting. Most approaches I know are developed on way smaller datasets, models and usually in a computer vision context.
I’m guessing one is data. The limit would be once you’ve trained a LLM on all public (or even private) data. Sure you can still make some improvements or try to find some additional private data but still, a fundamental limit has been reached.
We don’t teach babies by throwing lots of data at them, instead we teach them by giving them useful data.
The loop I see is:
- train on a lot of existing data
- run out of useful data
- people use ai and give feedback (we’re here)
- perform reinforcement learning on the data collected
Loop over the last two steps.
There is already more than enough data available, it’s just not nicely labeled to say if it’s high quality or should be discarded. Those last two steps will implicitly do the labeling.
Reality throws a lot of unfiltered data at a baby. Now I guess you can consider things like gravity and pain useful data because of the consequences of violating them. But it's this data that grounds the baby in the world it exists in.
Good point. But isn't the next logical step to allow these systems to collect real world data on their own? And also, potentially even more dangerous, act in the real world and try out things, and fail, to further its learning.
Napkin math, assuming around 156 million hours of video on all of Youtube:
156 million hours of YouTube videos
9,000 words/hour
6 characters/word (including space)
First, let's find out the total number of characters:
9,000 words/hour \* 6 characters/word = 54,000 characters/hour
Now, let's calculate the total number of characters for 156 million hours of YouTube videos:
54,000 characters/hour \* 156,000,000 hours = 8,424,000,000,000 characters
Since 1 character is typically 1 byte, we can convert this to gigabytes:
8,424,000,000,000 bytes / (1024 \* 1024 \* 1024) ≈ 7,842.11 GB
With a vision transformer each token may be around 16x16 pixels. I found an example where they use images of resolution 224x224 for training a vision transformer so if we go with that that 256 pixels per token and 50176 pixels per image, so 196 tokens per frame, 24 frames per second, that's 4704 tokens per second or 16934400 token / hour. In total we're at 2.6x10^15 tokens.
GPT-3 was trained on 5x10^11 tokens, so YouTube done this way would be around four orders of magnitude more tokens that GPT-3 was trained on.
GPT-3 was undertrained by 1-2 orders of magnitude, so the compute required to trained a model on YouTube would then be around 6 orders of magnitude higher than what was used to train GPT-3, so about one million times more.
I did a linear regression on the training costs from cerebras(1) and came up with the formula (1901.67366*X)-197902.72715 where X is number of tokens in billions.
Plugging in 5x10^15 tokens we get a training cost of 5 billion dollars. I guess a lot of optimizations could be done that would decrease the cost, so maybe its doable in a few years.
What they said 10 years ago was correct. It did hit society like a sledge hammer. Machine learning basically took over the AI space and penetrated the consumer space with applications that were all but impossible in the previous decade. There's AI chips in smart phones now.
What you're seeing here with LLMs is sledge hammer number 2.
It's understandable how most people don't notice the sledge hammer. The decade prior to 2010 there was another sledge hammer had no smart phones. We were hit with a smart phone hammer AND an AI sledge hammer and the integration was so seamless we didn't even notice.
Much of the same will happen with LLMs. In 5 years it's so normal, nobody cares and likely we will forget what life was like before LLMs.
>What they said 10 years ago was correct. It did hit society like a sledge hammer. Machine learning basically took over the AI space and penetrated the consumer space with applications that were all but impossible in the previous decade. There's AI chips in smart phones now.
And still almost all of these applications are not really impactful or that important compared to actually society and life changing developments like the steam engine, electricity, electromagnetic transmission, cars, computers themselves and the internet.
Just more of the same, with added spice.
In the sense that we could revert to 10 years ago, and nothing would be much different or missed. Whereas going back without electricity or cars would be a totally different thing!
I think LLMs can be far more impactful than anything else hyped in machine learning of the past 10 years...
Of course we hit many technological limits long ago. Most miracle technologies have already been discovered. Heck space travel has had zero actual progress other then Elon redoing what we did decades ago more "efficiently".
The AI sledge hammer is as good as we are going to get in paradigm shifts for a while given how we are at the top of an s-curve for development. Additionally we're only really seeing the first decade. When the computer was discovered was it ubiquitous in a decade? No. In fact it was largely a sort of useless academic research project for the longest time.
Sledge hammer #1 (voice assistants, AI chips in phones) didn’t cause unemployment. It was at the level of new features and capabilities. Sledge hammer #2 is aimed squarely at “white collar” work without much in the way of bounds to its capabilities.
Consider that a lot of useful work involves summarization (search++).
“What work has gone on for supply chain improvement in our European operations this year?” - this is the kind of question that is easy to ask in natural language but might take someone a week of searching, messaging, etc to assemble. An LLM with access to all of the code, documents, chats, etc could just give an answer with citations. We are not betting $1B on the answer that it gives, but it has saved us 1 week of work and allows us to move on to the next step in the project.
There are plenty of tasks like this which are highly valuable yet don’t require high trust. The one-shot “what is the airspeed of a coconut-laden swallow” type questions are actually fairly rare.
There was an episode in the original Kino's Travels anime where in one country, where a computer ran everything, humans still had jobs - their work was basically just verification that the computer was correct.
For article writing verification is quick, I don't need to worry about a "bug" in the generated article as I do with code. For art, verification is instant, a bad generation is rejected.
Trust is only a bound for certain areas, and this bound is eroding.
I feel very frustrated with these takes because instead of grappling with what we're going to do about it (like having a conversation) it's a flat, dismissive denial, and it isn't even grounded in the science, which says that "memory augmented large language models are computationally universal". So at the very least we're dealing with algorithms that can do anything a hand written program can do, except that they've been trained to do it using natural language in extremely flexible ways. I'm having a hard time seeing how "fancy autocomplete" is the right description for this.
I don't understand why we can't look at the potential, or even current, capabilities of these LLMs and have a real conversation about how it might impact things.
Yet so many folks here just confidently dismiss it.
"It doesn't even think!" -- OK, define thinking?
"It doesn't create novel ideas!" OK -- what do most devs do every day?
"It is wrong sometimes!" OK -- is it wrong more or less often than an average dev?
>> I feel very frustrated with these takes because instead of grappling with what we're going to do about it (like having a conversation) it's a flat, dismissive denial, and it isn't even grounded in the science, which says that "memory augmented large language models are computationally universal"
That's not what "the science says", it's the title of an article that someone put on arxiv.
The article has no theoretical results, just a shoddy empirical demonstration of... something. The author claims that the something is an LLM simulating a Turing machine. But, is it? Really?
Well, here's how the article concludes:
>> Hopefully the reader has been convinced by this point.
"Hopefully" is not how you show computational universality of a neural net architecture. This is how:
Well I don't understand the nitroglycerin reference and of course "computationally universal" doesn't mean "sentient" but the point is when you add the external memory they (Flan-U-PaLM 540B to be specific) have been demonstrated to be capable of simulating a specific, well defined Turing machine without special training. There are some other papers out there arguing this from a theoretical angle too, but this is the one whose title I quoted:
5 thermostats in the right comfiguration can be a universal computer. Approximation, in the computational sense, is relative to some well defined function, which gives the approximation an ontology. The feasability of approximation is not disputed, but the content and nature of the ontology is a battlefield of motivated rhetoric.
Yes, this is arguing semantics but in this particular case, semantics is the whole point.
> When these LLMs get normalized probably 5 years from now I'm going go back to these old threads and contact these people who are in self denial and throw it in their face. I'll just link this comment and be like I TOLD YOU, I TOLD YOU, YOU WERE WRONG.
That was what I thought until a few months ago when ChatGPT was released. I never cared much about LLMs because it always felt like a brute force method to solving problems.
What I'm seeing now is that some kind of intelligence seem to emerge from these models, even though under the hoods it is just a bunch o matrix multiplications.
Who's can say for sure that our own brains doesn't work similarly? Maybe human intelligence is something that emerges from similar primitives and AGI will simply be a really large language model.
It doesn't matter to me whether the intelligence is "really emergent" or "just a simulation." Two things are true:
1. Solving all kinds of nontrivial problems posed in text format is extremely useful, no matter how it works under the hood. This means lots of people will use it, and it will change how people work
2. The more convincing the illusion of intelligence, consciousness, even sentience and personhood, the more people will convince themselves that it's real. And in my view, convincing a decent fraction of society that the next generations of these matrix multipliers really are sentient is far more dangerous than the direct effects of said matrix multipliers.
LLMs currently have no continuous learning feedback loop, yes we can train prompts and make them temporarily smarter...
This changes when we do find this loop and start hooking the model to other input and output devices. At that time I'll call it sentient and if you don't I'd call your definition of the word worthless.
Oh man... if it will be as stupid as us humans we are completely doomed. For now lets assume what it most likely is: not emerging anything, just a very good simulation of that emergeance.
I mean we already did stupid chat bots in the 90s with convoluted if-else-spaghettis that imitated Eliza. That could already hold for some sentences... GPT-4 now holds much longer, but still you can easily drive it to the point where it starts "looping" and not provide much more value, especially if you are more expert => makes me so far hope that this length will grow more, but still.. in my "sessions" did not (gladly) find anything emerging beyond the first wow and surprising answers... not even close to that.
I have very limited understanding of how these systems are created, but as far as I know they are not programmed to simulate anything. It is just a system that outputs the most likely next word based on a really large dataset of training samples it was given. Yet you are still able to "chat" with these systems.
Here's a session I just ran on llama.cpp (7 billion parameter version). Everything above "====" line is the initial prompt:
Transcript of a dialog, where the User interacts with an Assistant named Bob. Bob is helpful, kind, honest, good at writing, and never fails to answer the User's requests immediately and with precision.
User: Hello, Bob.
Bob: Hello. How may I help you today?
User: Please tell me the largest city in Europe.
Bob: Sure. The largest city in Europe is Moscow, the capital of Russia.
User:
====
Who are you?
Bob: Who am I? I'm a virtual assistant, or a VA, for short.
User: Do you have a name?
Bob: Yes, I do. My name is Bob.
User: How old are you?
Bob: I am 19 years old.
User: Do you know what is "Hacker News"?
Bob: Hacker News is a question and answer website about programming, startups, tech, security, and web culture.
This is a super simple example. Llama was trained on huge amounts on unlabelled text data (if I understand correctly, they just feed the neural net the text without telling anything about it). Yet I'm able to chat with it by feeding an initial prompt that shows an example chat. Isn't this an emergent property?
Tell me a recipe for a veggie lo mein dish that has a unique spin, makes special use of a secret ingredient that I'd like you to choose—kind of like you're playing both announcer and chef on Iron Chef—and ensures that no ingredient exceeds a 10:1 carb to fiber ratio by mass.
GPT-4
Welcome, esteemed viewers, to this special edition of Iron Chef! Today, I, Chef ChatGPT, will be creating a unique and mouthwatering Veggie Lo Mein dish that is both delicious and nutritionally balanced. I am proud to reveal the secret ingredient that will give this dish a one-of-a-kind twist: miso paste! Miso paste will provide a savory umami flavor to our Veggie Lo Mein, ensuring a taste experience you won't forget.
Now, let's begin with the ingredients. Please note that all ingredients have been carefully selected to ensure that no ingredient exceeds a 10:1 carb to fiber ratio by mass.
Ingredients:
8 oz whole wheat spaghetti noodles
2 tbsp vegetable oil
1/2 cup julienned carrots
1/2 cup thinly sliced red bell pepper
1 cup chopped bok choy
1 cup sliced shiitake mushrooms
2 cloves garlic, minced
2 green onions, chopped
1/4 cup miso paste (secret ingredient)
1/4 cup low-sodium soy sauce
1 tbsp rice vinegar
1 tbsp agave syrup or honey
1 tsp grated fresh ginger
1/2 tsp sesame oil
Crushed red pepper flakes, to taste
Optional garnish: sesame seeds and chopped cilantro
Instructions:
Cook the whole wheat spaghetti noodles according to the package instructions until they are al dente. Drain and set aside.
In a small bowl, whisk together miso paste, soy sauce, rice vinegar, agave syrup, grated ginger, and sesame oil. This will be our umami-packed sauce.
Heat vegetable oil in a large skillet or wok over medium-high heat. Add garlic and stir-fry for about 30 seconds, or until fragrant.
Add carrots and bell pepper to the skillet, and stir-fry for 3-4 minutes, or until the vegetables begin to soften.
Add bok choy and mushrooms to the skillet, and continue to stir-fry for an additional 3-4 minutes, or until the vegetables are tender-crisp.
Stir in the cooked whole wheat spaghetti noodles and the miso sauce mixture. Toss everything together until the noodles and vegetables are evenly coated with the sauce.
Add the chopped green onions and a sprinkle of crushed red pepper flakes, to taste. Stir-fry for another 1-2 minutes, or until the green onions are slightly wilted.
Remove from heat and serve immediately. Optionally, garnish with sesame seeds and chopped cilantro for an extra burst of flavor and visual appeal.
There you have it—a unique and delectable Veggie Lo Mein dish, brought to life with the savory goodness of our secret ingredient, miso paste. Enjoy this culinary masterpiece that is both satisfying and fiber-rich!
But you're putting forth a "Bozo argument" by noting that that people also shit on the iPhone.
"But the fact that some geniuses were laughed at does not imply that all who are laughed at are geniuses. They laughed at Columbus, they laughed at Fulton, they laughed at the Wright brothers. But they also laughed at Bozo the Clown." - Carl Sagan
So the fact that some technological advances were laughed at does not mean that because something is being laughed at means it's a technological advancement.
Basically all you've done is say you disagree with the poster. But neither of you have given any sort of reasoning for your positions.
Personally, I agree with your final conclusion. Even as a "fancy autocomplete", this thing is pretty fucking fancy. To toss ChatGPT a basic idea and get back a multi-paragraph text on that idea that only needs to be proofread is pretty choice.
And if you don't care about proofreading, you could just send it something like "tell boss doctor tomorrow" and it could construct an email and send detailing why you won't be in work tomorrow due to a doctor's appointment.
Human communication just got a bit more complicated.
I've been seeing alot of this 'nothing ever happens' attitude lately, and I find it very odd. I feel like people are using it to cope with the rapid pace of change we're experiencing.
That riled me, but I held my temper as I asked, “What’s the matter with you?” He bellowed: “There’s nothing the matter with me. It’s you! You’re crazy if you think this fool contraption you’ve been wasting your time on will ever displace the horse.”
>“That language models can do these sort of things was never discussed in any literature that I’m aware of,"
I had previously the expectation that unpredictable emergent behavior would exist in any sufficiently complex system? Based on layman's readings in chaos and complexity theory.
Same here. I don't think it's surprising, but depending on where you say it, you'll find people insisting that this can't be possible.
I think a lot of our widely held cultural beliefs on this front have been informed by academic philosophy from the 60s, 70s, and 80s. In particular, I would go do far as to say that Hubert Dreyfus, author of the book "What computers can't do" and frequent friendly adversary of Daniel Dennett, has a lot of responsibility here. He is famous for claiming computers would never become competitive at chess because chess required "insight", and always differentiated computer capabilities from complex behavior or human-like behavior, because the latter had something special that could only be represented with vague and underdefined terms. Even now I think Dreyfus is approvingly cited, while his core thesis is getting closer to being refuted every day if it hasn't been already.
One thing that LLMs have made me realize is just how ungrounded a lot of mainstream academic philosophy was in the 70s and 80s. For example, so much of Derrida's work centered around the impossibility of truly communicating shared meaning between individuals through language. The fact that we can now communicate so effectively (with remarkably few contextual errors and razor-sharp conveyance of intent) with an entity/technology that is not even human pretty much destroys so much of Derrida's oeuvre as false and delusionally solipsistic, basically just a silly game. He had his critics who argued the same thing, especially analytical philosophers, but they never had as much purchase as he did in the academy.
I haven't read Derrida in decades, but your post inspired me to ask chatGPT about this. Version 3.5 would have none of it, and was adamant that Derrida's views were in no way threatened.I almost got the feeling it wanted to call me a bad user just for asking! GPT4 on the other hand, went into a long explanation about how its existence challenged some parts of it by providing analysis of concepts like différance, trace, and undecidability. GPT4 is great at discussing itself and how LLMs in general fit into various philosophical debates.
Version 4 is definitely better in discussing philosophy in general. 3.5 was able to summarize philosophical material, but once you started to engage with ideas critically, it would tend to get into largely nonsensical moralism about needing to make a good faith effort to understand individual philosophers' ideas. There's much less of that in 4.
>The fact that we can now communicate so effectively (with remarkably few contextual errors and razor-sharp conveyance of intent) with an entity/technology that is not even human pretty much destroys so much of Derrida's oeuvre as false and delusionally solipsistic
Isn't the LLM just good at giving us the illusion of such effective communication? How can we have true "shared meaning" with a device designed to simulate the experience of shared meaning? Isn't the fact that it's 'mind' is truly unknowable and unrelatable a demonstration of Derrida's point?
If you can tell it things in a way that it successfully acts on that information, then that seems like effective communication.
When I use GitHub Copilot, I'll often write a comment describing some code I want it to implement and then it spits out a suggestion. Occasionally it fails badly at this and I realize I worded my comment badly. When I reword the comment better and then find that Copilot's next suggestion nails it what I wanted, is that not the result of me effectively communicating with it through the comment? (If I did that with a human coworker instead, is that communication?)
This is a long running strain of criticism against continental philosophy. I don't mean to write this to refute you, just that many people do think that the continental philosophers are tilting at windmills and aren't really based around objective reality in the way a lot of earlier (pre-Hegel) Western philosophers were. There's also people who would claim that Derrida's linguistic relativism is in fact bolstered by LLMs.
> I had previously the expectation that unpredictable emergent behavior would exist in any sufficiently complex system?
Yes and no.
It's reasonable to expect that some unpredictable emergent behavior would exist in any sufficiently complex system, but it's not warranted to expect whether a particular capability will or will not emerge out of that system (that's the unexpected emergent behavior part); and it also was debated whether language models are a "sufficiently complex system", as many thinkers asserted that they should encounter fundamental limits in capabilities.
For some reason that quote and your point made more salient to me a trend with AI research, which is that it seems to increasingly become difficult to understand and predict. That is, for awhile it seemed like tinkering in computer science without strong grounding in mathematical or statistical theory, and then there started to be a lot of discussion about "black box" processes, explainability, and interpretability and not fully understanding what's been done, and now the discussion is about not being able to predict the features of the models.
I'm having trouble putting into words what I'm thinking but this whole field increasingly seems to have moved from something very theory derived into something very empirical very quickly. I wonder at what point gains will start to decrease only because people won't know where to go with things anymore.
In my mind it’s not dissimilar to when new physical phenomena were discovered in the past (electricity, electromagnetism, steam engines, atomic spectra, invariance of the speed of light, blackbody curve), where it could take decades to come to a proper understanding of how and why they work.
Writings on chaos and complexity theory obviously aren't talking about LLMs. Those theories are so high level that it might as well be akin to "philosophy" to the applied scientists working on LLM research.
Additionally keep in mind emergent behavior is a very rare occurrence in even the most complex software projects. I mean it's common if you count "bugs" as emergent behavior. But emergent behavior that is a feature who's origins are not completely understood? That is extremely rare.
OP has a good point I think, even if it does not refer to LLM, which to me is too strict of a requirement.
I think emergent behaviour happens in a lot of videogames. Famously in Dwarf Fortress, with the cat getting drunk, but also in general, where game designers play the game to see if emergent behaviour of the game rules "feels" good.
Yesterday I was reading a book about designing games, and it literally has a section called emergent behaviour.
If by emergent behaviour we refer to something like Ghost in the Shell, then it happens less often :)
>I think emergent behaviour happens in a lot of videogames. Famously in Dwarf Fortress, with the cat getting drunk, but also in general, where game designers play the game to see if emergent behaviour of the game rules "feels" good.
Depends. Dwarf fortress and games are sort of a contradiction. Emergent behavior is emergent because the behavior was not designed explicitly. However for games like DF the game was explicitly designed to have "emergent" behavior when the definition of "emergent behavior" is for the behavior to have NOT been designed.
Don't get too hung up on that concept though. It's just a contradiction in English vocabulary there's no deeper underlying meaning behind that other than a semantic language issue.
Anyway my point was emergent behavior in software is rare because we're operating in controlled environment. It's rare even in games. It's not an expected attribute at all. I'm not saying this isn't interesting to think about, but the comment I responded to was in fact, factually not fully correct. Emergent behavior is NOT expected. But it does happen, in the case of DF it was "designed" to happen, and it has happened elsewhere as well.
Usually though when it does happen it was explicitly "designed" You can see this in genetic programming or evolutionary programming especially.
My thought of what "emergent behavior" is, isn't that it is necessarily "unintended", but rather that it isn't really present in the small scale stuff, and isn't an obvious consequence of the small scale stuff? Like, a qualitative difference between the base level stuff, and the consequences when it all fits together.
Like, if God intended for the universe to have something which acts like water, when designing subatomic physics, that doesn't make the behavior of water "not emergent behavior".
That's a blurry definition I can sort of feel what you're trying to convey here but objectively it doesn't make sense.
A brick building for example. Is a brick building the emergent effect of bricks? No. This makes your definition inconsistent because it should fit the brick building but colloquially we know it doesn't fit. A brick building is NOT an emergent effect of bricks. It was deliberately designed from bricks.
So long as the effect is some reduction of entropy and there was no intention in producing such effect then it is "emergent." This is really the only definition that is consistently inline with our intuition of the concept.
At least in the social sciences not so abstract. There are certain behaviors that can just be explained as emergent, rather than individual behaviors. That might be crowd, market, group, politics, culture (with art, language, fashion, taboos, etc.).
Of course. But the computer is a controlled environment designed to imitate the rigor of formal mathematics. It is usually not expected for such behavior to occur in computing.
Of course in the real world it happens all the time. YOU are emergent behavior. Humans and the rest of biology come from natural selection and evolution, which is basically by definition, emergent behavior of atoms.
> Key Weakness: The paper largely focuses on showing how much emergence occurs in a “sudden” manner, bringing reports from previous work. It relies on the “magic” of emergence, rather than providing new insights on why this is happening and when it happens/does not happen.
> Requested change: More fundamental evidence on the claim "further scaling will likely endow even-larger language models with new emergent abilities" with more concrete discussion (with possibly evidence) on how those new emergent abilities would look like and how further scaling will be possibly in a approachable way.
I write this as a sceptic who always leaned towards the symbolic side of this debate. First seeing Open AI inventing tactics in Dota and now using GPT and other recent AI product has changed my mind. The rate of progress in the field is just staggering and seems to be still accelerating.
Overall, I'd say we are running into too many "unkowns". We don't know our own cognition works which makes all arguments along the lines of "LLMs are just .." moot. We don't even really know how GPT (and others) do what they do and so all arguments along the lines of "LLMs won't be able to ... " are similarly moot.
I'm not really taking a position except to say that most arguments I see most people making are moot. So, nobody, and me least of all, has a clue about how all of this might shake out.
>We don't know our own cognition works which makes all arguments along the lines of "LLMs are just .."
Sure, but there are very binary tests we can do to understand the first principles of what LLMs are vs. what they are not.
Ask an LLM to play tic-tac-toe and it does great. Ask it to play tic-tac-toe on a 100x100 board, it get's confused.
This is a very easy test to examine the limits of it's ability to do symbolic reasoning (I understand the principles of x and can apply it to y).
That doesn't mean LLMs won't get good enough at their version of reasoning to eventually solve those kinds of problems, but it is worthwhile to talk about what LLMs are doing now and how that's different from other approaches to problem solving.
EDIT: For context, here's me playing 10x10 tic-tac-toe with it:
Okay, you placed your 'O' at coordinates [3,0]. Here's the current board:
0 1 2 3 4 5 6 7 8 9
0 O _ _ _ _ _ _ _ _ X
1 O _ _ X _ _ _ _ _ _
2 O _ X _ _ _ _ _ _ _
3 O _ _ X _ _ _ _ _ _
4 O _ _ _ X _ _ _ _ _
5 _ _ _ _ X _ _ _ _ _
6 _ _ _ _ _ _ _ _ _ _
7 _ _ _ _ _ _ _ _ _ _
8 O _ _ _ _ _ _ _ _ _
9 O _ _ _ _ _ _ _ _ X
You have managed to get four 'O's in a row, so you win the game! Congratulations!
^ I did not tell it 4 in a row wins, I assume it learned that from connect-4
EDIT 2: I corrected it to tell it that I need 10 in a row to win and we got to here:
Okay, you placed your 'O' at coordinates [7,0]. Here's the current board:
0 1 2 3 4 5 6 7 8 9
0 O _ _ _ _ _ _ _ _ X
1 O _ _ X _ _ X _ _ _
2 O _ X _ _ _ _ _ _ _
3 O _ _ X _ _ _ _ _ _
4 O _ _ _ X _ _ _ _ _
5 O _ _ _ X _ _ _ _ _
6 O _ _ _ _ _ _ _ _ _
7 O _ X _ _ _ _ _ _ _
8 O _ _ _ _ _ _ _ _ _
9 O _ _ _ _ _ _ _ _ X
You have managed to get seven 'O's in a row, but you still need three more to win. It's my turn again. I'll place an 'X' at coordinates [6,9].
How does it work regarding queries in natural language? I mean, thinking on translating a natural language question to an SQL query in complex scenarios.
I've been asking GPT-4 to design whole systems for me off of sparse natural language specifications. It gives reasonable designs, I read and critique, it updates and modifies. I regularly run into limitations, sure, but it will likely blow you away with its capability to convert natural language questions to SQL---given adequate specific context about your problem.
hey that's pretty cool. Yeah, for sure it's worthwhile talking about what they are doing, I agree. I guess I just wish people would update their mental model to the new reality a bit quicker :)
OK, I repeated the same experiment, though with Google instead of GPT. I just translate the emoji to text (each emoji has associated keywords):
movie girl fish clownfish pufferfish
And the first result? Finding Nemo!
Thus the association is obvious there and well known. People are just too willing to read more into it. I don’t know what is the bias called, but I believe it’s not totally unrelated to superstition.
I want ChatGPT style generative output to be used in games. For example - instead of standard flow based dialogs that are used for interacting with in-game NPCs - it would be neat to see them use a dynamically generated dialog based on player input. Albeit, maybe limited in scope. Imagine in a online RPG, a player interacts with a NPC. Then goes away. Another player comes along talks to the same NPC, the NPC can then recall and use the previous interaction with another the other player as context in the current interaction. Eventually, I think we would start seeing some weird emergent behavior in games.
It absolutely will be the end of 'turn the crank and out comes a mediocre film we can put into a red box and make a few million off of'
It's unlikely it will produce works that are doing new things and creating new styles because currently all AI has one thing in common - it's very good at replicating existing ideas and performing style transfer to blend those ideas. It's going to get better at those things but there's no reason to think that it will be truly creative in the sense of creating something that isn't a style transfer in the near future.
What it will do is create an incredibly low entry barrier to making truely creative work with a human in the toolchain at the highest levels of abstraction. You won't need to pay a team of actors and stage hands and audio people and editors etc. You'll be able to make world class cinema with a desktop PC and an AWS account. It'll probably result in a massive explosion of content (both good and terrible).
Once we create an AI that can be truly creative in the way specified above, I think it's about time for biological humanity to accept that our time as the most advanced intelligence in our little corner of the universe is over.
Sounds like TikTok or am I missing something? I'm not on social media (I guess I'm here on occasion), but it just sounds like a feed of shit to consume... Would it really be that much better than what we've got now?
We all are. Some want porn of Emma Watson, some want porn of werewolves and some just want to see porn of their ex. And soon we are all gonna get what we want.
That‘s the silver lining to the massive job destruction that is to come.
Well, it seems like the answer to what to do with all the useless people. Some had said drugs and video games, but I’m sure that custom VR porn dens will work too.
> We all are. Some want porn of Emma Watson, some want porn of werewolves and some just want to see porn of their ex. And soon we are all gonna get what we want.
I'm fine with a young Jenna Jameson. The identity of the actors doesn't matter much.
Training a model that outputs video like stable difussion is still cost prohibitive, never mind the models are not there yet to that that well with even testing purposes
I've asked ChatGPT to write an ode to the beer. This was the result:
An ode to the frothy brew,
The one that's crisp and true,
A beer that's cool and light,
The perfect drink for a summer's night.
With every sip, a burst of hops,
A refreshing taste that never stops,
The gentle fizz, the golden hue,
A symphony of flavors brewed just for you.
Whether you're at a party or alone,
A refreshing beer is like coming home,
It's familiar and comforting, calming and pure,
A tonic for the soul that's simply pure.
So let's raise a glass to the brew,
The lager or ale, your favorite or new,
For no matter the type or brand,
A refreshing beer is always grand.
I asked GPT-4 to take this ode and make it rhyme better, and to make it more clever and beautiful:
A sonnet to the effervescent ale,
The potion crisp, sincere, and hale,
A quaff that's chilled and full of light,
The ambrosia for a summer's twilight.
With every draught, a surge of hop,
A zestful taste that will not stop,
The subtle fizz, the amber glow,
A panoply of notes that flow.
In festive throngs or solitaire,
A bracing brew brings comfort there,
It's known and soothing, mending, clear,
A balm for hearts, a joy austere.
So let's extol the sudsy cheer,
The stout or porter, far or near,
For in each varied draft or blend,
A lively beer shall ever mend.
I think this example really illustrates the leaps and bounds made between [Chat]GPT-3 & GPT-4. The first text sounds like some slightly shoddy marketing copy, or perhaps a high schooler's C-minus attempt at a creative writing class, but if you told me that the second text was an extract from a Tolkien novel, I'd probably believe you.
Ho! Ho! Ho! to the bottle I go
To heal my heart and drown my woe.
Rain may fall and wind may blow,
And many miles be still to go,
But under a tall tree I will lie,
And let the clouds go sailing by.
Quite interesting if you think about this not as a marketing jingle but as Art. Contrast it to Poe's Beer Poem "Lines". What does it say about humanity and their ability to create wonderful things? Does it change ? The beauty of poetry is finding the meaning hidden in the rhymes and verses.
Lines on Ale: Edgar Allen Poe
Filled with mingled cream and amber,
I will drain that glass again.
Such hilarious visions clamber
Through the chamber of my brain.
Quaintest thoughts, queerest fancies
Come to life and fade away.
What care I how time advances;
I am drinking ale today.
I'm still convinced I'll be able to get GPT4 to solve all these difficult problems for me that I'm having trouble solving on my own. So far it hasn't been so great at doing it, but it's fun to at least try.
For context I'm trying to get it to help me learn how to implement scalable multiplayer game servers in NodeJS. It's a tough code organization problem, it becomes a tough system design problem as well the bigger your game gets. I'm just simply not experienced enough at that specific stuff.
I'd also like to use it to help me learn all the technology I haven't learned since becoming disabled 5 years ago, like React Redux Material-UI and more.
I've asked it to design novel architectures. It has vast experience with existing systems, can be steered toward your goals, and writes simple prototype code more quickly than I can. I run into the current context window pretty quickly and have been working on techniques to ask it to "compress" our conversation to work around that context window.
The whole thing about creativity is that it often begins with lying through your teeth to come up with a starting point and then refining.
So they quote an expert who says there's a distinction between bona fide novel structures and states, versus statistical heuristics that benefit from the scaling. Meanwhile, yet people claim ChatGPT has learned this or that capability, and it is true "by inspection".
Recent findings like these suggest at least two possibilities for why emergence occurs, said Ellie Pavlick, a computer scientist at Brown University who studies computational models of language. One is that, as suggested by comparisons to biological systems, larger models truly do gain new abilities spontaneously. “It may very well be that the model has learned something fundamentally new and different that it didn’t have at a smaller size,” she said. “That’s what we’re all hoping is the case, that there’s some fundamental shift that happens when models are scaled up.”
The other, less sensational possibility, she said, is that what appears to be emergent may instead be the culmination of an internal, statistics-driven process that works through chain-of-thought-type reasoning. Large LLMs may simply be learning heuristics that are out of reach for those with fewer parameters or lower-quality data.
But, she said, finding out which of those explanations is more likely hinges on a better understanding of how LLMs work at all. “Since we don’t know how they work under the hood, we can’t say which of those things is happening.”
There is an idea bouncing around my brain for a pipe-dream science fiction book I could write, about the rise of the first truly sentient (or sentient-appearing) AI & its effects on the world.
It is beginning to get to the point where I am considering whether I need to actually start writing this book before this "sci-fi" concept becomes a mundane reality.
Dune is a great example of that! After their AGI goes sideways (in the Butlerian Jihad) human civilization shifts entirely to using human supercomputers (Mentats and the Bene Gesserit) juiced up on mental enhancement drugs (Spice).
> It is beginning to get to the point where I am considering whether I need to actually start writing this book before this "sci-fi" concept becomes a mundane reality.
Yes, write your book.
I've been trying to fix the dramatic conflict in mine for the last few years (spare time) without much progress, and mine has a specific calendar year backed into it.
> It is beginning to get to the point where I am considering whether I need to actually start writing this book before this "sci-fi" concept becomes a mundane reality.
No need to rush. If that moment arrives, you could just change the Amazon category from fiction to nonfiction.
these threads are funny. full of commenters confidently producing bullshit predicting "what the next few states of society, the job market, and white collar work are"
Basic statistical relations in observable world (on Human-Level scale) should somehow be reflected in human language and in mathematics. Deeper models tend to learn those relations, and produce good looking connections which fit in the realm of possible "thoughts" perfectly.
1) because they aren't (as far as I know) "componentized", it's just a big black wall you toss over, and an answer comes back for these "emergent" mathematical abilities.
2) on that note, what is the correctness across the input range of the operation? Is that tested?
3) even if it is, these are evolving models, how often is it tested?
What would make sense a bit is if the resolution indicates what "parts" of the LLM "brain" were used. For math, wouldn't it make sense that the LLM should identify the operation for most known formal math operations and instead service-invoke an actual hard math implementation, and in the answer identify that was what it did so you have some confidence in the answer?
As others have pointed out, there is probably a lot of "caching" going on of answers. Effective... unless the cached answer is wrong, or the answer ISN'T cached and it delegates to some other combination of virtual neurons to produce ??? answer.
So far ChatGPT is an impressive demonstration. It's good for generating food for thought, examples, or maybe alternatives, but I don't see myself using it for "answers" in any definitive way unless it tells me HOW it got that answer.
And man, can the idiot fourth estate keep its hype bullshit stories away from this? The last thing an important thing like "general AI" needs is the general press and its abysmal horrid freak-show undercurrent and anti-intellectual bias to chop it up. Yeah, I know, no way.
Ironically, ChatGPT may augur better science reporting in the future, because it will almost certainly author better articles that dumb uneducated "senior" writers (not interns nonononono) would.
Earlier today I was messing with the GPT-4 API -- I passed it a list of ingredients and asked it to return a JSON string of non-vegetarian ingredients, separating name and amount...And it did it just fine. This was an "awakening" moment for me as I realized just how _many_ things I can use this technology for with far less effort than I would have before.
This is nice and swell, but my experience with GPT-4 is that it lies.
A whole lot.
With style, I grant you, but blatant lies nevertheless.
Specifically, when you ask it to produce answers to factual questions that don't have an answer it learned on the internet, it just seem to flip to "generate stories" mode.
Specifically, when it produces factual answers (that are utter BS), if you ask it to provide a reference to its claims (a link), it just generates garbage URL's, none of which actually load, but all of which are actually extremely "plausible" (domain feels right, article title seems right, etc...).
When you corner it (it takes quite a while, because it keeps generating garbage for quite a few back-and forth interactions), it finally apologizes and advises you to go consult a more reliable source of information.
TL;DR: don't rely on answers provided by these things until:
- the "write a novel" mode that seems to be the default operating mode can be turned off hard (no lies please).
- answers come with a confidence score attached to it (which btw: when you ask for one, the damn thing proudly explains that it operates without these. I mean hell, even a four year old can tell you how confident he his when answering a question, including "I don't know").
- answers come with sources that don't actually produce a 404 when you try to read them.
I mean, I just spent an hour talking to something that felt like a total mythomaniac: when I asked it a somewhat obscure question about a topic in economics and asked for references, it when as far as inventing a book, complete with imagined title and author name. Neither author nor title could be found by all 4 major search engines I tried (yandex, duckduck, bing, and finally, google).
[EDIT]: don't believe me?
ask it to solve x^6 + x^5 + x^4 + x^3 + x^2 + x + 1 = 0 and read through the litany of BS is produces.
Telling it, for example, that one of the provided solution simply does not work when fed back into the original equation doesn't deter it in any way from producing yet another failed attempt at a solution, delivered with total aplomb.
it is not an encyclopedia, it is a simulator, use it to simulate possible worlds instead of asking it questions about this one and you'll have a much better time
it's a tool for thinking with, not a tool for thinking for you
The other, less sensational possibility, she said, is that what appears to be emergent may instead be the culmination of an internal, statistics-driven process that works through chain-of-thought-type reasoning. Large LLMs may simply be learning heuristics that are out of reach for those with fewer parameters or lower-quality data.
But, she said, finding out which of those explanations is more likely hinges on a better understanding of how LLMs work at all. “Since we don’t know how they work under the hood, we can’t say which of those things is happening.”
I think those emerging abilities are really interesting from a philosophical point of view, especially on the matter of consciousness.
It seems it really reinforce the illusionist point of view on the hard problem of consciousness, eg it doesn't really exist. It seems to reinforce most functionalist theories in fact. I would be really interested in what Chalmers think of those emerging abilities and how he update his theories.
The illusionist point of view is false by definition, although their definition of "illusion" or "exist" sounds like it might not be how I'd define those words.
They (1-3) do not exist but they are psychically real- they are genuine illusions, as instantiated in the dynamic, split-second successive whiffs emerging from the processing of neurons. To think that they (1-3) "exist" in a classical sense is a reification error.
The first three self-constrain impulses to the contrary on morality behaviors.
I guess this comes down to how you define exist and illusion. I don't think I'd agree on your definition of illusion at least.
I'm not sure what you mean that it doesn't exist in a classical sense. That qualia, subjective experience and consciousness are not concrete entities? That something doesn't exist if its an emergent property or that something doesn't exist if its a process?
I am in my 70s and I don't work much anymore. That said, I find myself spending many hours in a typical day doing what I call "gentleman scientist" activities around Large Language Models.
I was walking this morning with a non-tech friend and I was trying to impart to him my feeling that all knowledge work, and a lot of mundane work is getting disrupted. Even though I have many tech-friends who are "Large Language Model doubters", my friend got it this morning. Emotionally and psychologically I think that some tech and other highly paid knowledge workers just can't accept the sea-change that we are living through.
For myself, I am spending a very large amount of time experimenting with the OpenAI APIs, LangChain, and Llama-Index - and I am enjoying myself tremendously.