Please name me one technological advance of major import in the fundamental transformer kernel space that has occurred in the last decade that has any import at all on today's LLMs.
it is unlikely that the output of LLMs will improve.
there is no fundamental breakthrough in transformer technology (or anything else) powering todays' LLM "revolution"
There is only scale being employed like never before => vast datasets being plowed through being sufficient to provide the current illusion for the less observant humans out there...
10 years from now this current fad of LLM's pretending to be intelligent will look preposterous and unbelievable: "how COULD they all have fallen for such hype, and what a cost of joules/computation... the least deterministic means possible of coming to any result... just wasteful for no purpose..."
Containerization and orchestration of containers vs learning how to configure HaProxy, how to use Certbot, hmmmm
The questions you pose are legit skills web developers need to have.
Nothing you mentioned is obviated by K8s or containerization.
"oh but you can get someone elses pre-configured image" uh huh... sure, you can also install malware. You will also need to one day maintain or configure the software running in them.
You may even need to address issues your running software causes.
You can't do that without mastering the software you are running!
Author presumes we need Docker in the first paragraph. I suppose this is why they think we need Kubernetes.
I propose using "the operating system" as the basic-unit here.
It already runs on shared hardware thanks to a hypervisor.
Operating systems know how to network.
The entire desire for Docker in a production app comes down to willful ignorance of how software one depends upon is configured.
I have no need for Redis in my life. There is nothing unique it provides in 2024, and they have no special sauce I would consider getting hooked-on (locked into).
I am trying to remember why their software became considered ubiquitous for caching and sessions, and I reckon many a framework is busy rectifying this choice, as we speak.
... but, garbage collection, no direct memory management = impossible to write any firmware etc that needed to directly set bitflags and such.
how could you call Go "low-level systems programming"?
Relative to Ruby, an interpreted dynamic language, sure, I will grant you that Go is lower-level, in the sense that you can compile a binary executable from it.
Go also happens to have web server primitives in their standard library and you COULD build a web app with only net/http...
I would say that bad ideas are having a heyday across all economic sectors, not only tech. You could call it the epoch of bad ideas having a heyday, if you want.
One has to be careful around opportunistic gold-rushes; if one is not actively purveying gold, one may be getting rushed.
Business as a whole is becoming much more unfriendly to labor. They're realizing that perceived external threats like temporary recessions, AI labor, a threatened US dollar, etc, can give them continuous leverage in any labor discussions and they're pressing the advantage for all that they can.
If the US sets a debt ceiling and inevitably hits it (based on past performance), the ensuing devaluation of the dollar may lead to an even stronger form of corporatism that makes such power plays about wages, hours worked, and work conditions look relatively tame.
Yup, that's part of the supposed "Gen Z is lazy" mentality they try to spread. While making them jump more hoops than ever for "entry level" jobs that has less buying power that may not even pay rent. And no pension/social secutity to even try to reel the few in that do that.
People may not all be financial experts, but they can do basic math and realize none of this makes sense.
>If the US sets a debt ceiling and inevitably hits it (based on past performance), the ensuing devaluation of the dollar may lead to an even stronger form of corporatism that makes such power plays about wages, hours worked, and work conditions look relatively tame.
2 years ago we were starting to properly talk about 32 hour workweeks. Sad how quickly things can go backwards. Really hope America isn't stupid and tries to go the Greece route with this inevitable crisis.
Then you tell someone to start their own business so they can treat their workers the way they deserve to be treated, then you get a bunch of excuses about why that is impossible! I guess there is just a conspiracy from every business to prevent honest people from being in charge
I work for a company which treats workers very well and the company is doing very well themselves. Of course it's possible; don't let anyone tell you it isn't.
How can I do it without capital, straight out of college, with only around 2 years of real-world experience and no network? Honest question.
The only play that occurs to me is surviving off of gig work, building the business as a sidehustle. But I've seen so many people who seem to be permanently stuck there, with no real business to grow and no way to explain their resume gap to employers.
From what I've seen lurking here, successful bootstrapped businesses come from experienced people who know what they're doing and have savings to fall back on.
Find a good local problem that you have connections with people who'd buy it.
Doesn't need to be a software, many small businesses start as sole proprietor working for himself and slowly growing the company by hiring help.
No cheap programmer would copy your "fixes-fences-in-Boston.com" idea. A lot of local services aren't sold properly on the internet, so if you combine the two you can get something out of your time and labour.
Also local bureaucrats love to "regulate" and automating local compliance is also a good niche. Now with all LLMs around the scope of what is possible has grown, thus the niches where it could be applied have grown too.
Don't drink the VC/YC combinator cool-aid, that you "go big or go home". It's better to own 100% of your comany, than 3% of a VC based startup most of the time. You see outliers like Facebook, Airbnb & etc... but as 37Signals has proved for the majority of startup founders the risk/reward ratio is skewed not in their favor.
There are many local problems that require local knowledge and serving local customers.
I have a friend with a business for California vehicle compliance reports. Some stupid paperwork that needs to be updated yearly when the rules change. It's super local and he has 10 employees supporting clients remote and on site. It grew very slowly but it's in 11th year now and revenue is not bad at all. Nothing to compete globally, knowledge is local, clients need local services.
What all indie hackers are doing is getting support and resources from an equivalent of venture capital investment. Which is Cathedral approved education that effectively reduces to attaching an epistemology onto yourself that limits you and prevents you from functioning outside the cramped divisions of civilization-approved entrepreneurship.
The education offered by civilization includes logic that's crafted by a pedagogy that's biased toward vulgarity and social skills that don't perform well when it comes to building anything that would help a man get away from a forced commitment to anything more than maximizing viewer impressions on provocative Internet-uploaded content. The average tech entrepreneur isn't any better than a McDonald's hamburger grill operator or a female OnlyFans model or a delivery app driver, once you remove your civilly trained bias toward low resolution videos on socioeconomic dynamics.
To use a great illustration, even a billionaire techno-commercialist like Elon Musk can never hope to achieve independent wealth acquisition. Because his education, personal origin, and development are not really deviant and an instance of someone that can perform beyond the abstract black box that is knowledge given by civilization's life experiences which the life offers to every man. He will always come up short, whenever it's a question of exiting from the liberal-democratic regime and its permanent ironic anti-libertarianism stance. The question is: How to build a business that's not required to conform with all expectations, social and physical/ontological? Leading up to aerospace technology and acquiring science for Earth-to-Mars orbit transfers ain't it. Even having a business starting capital of one hundred dollars in the style of the $100 Startup comes with a history whose financial system component is tied to having a certain social obligation. A certain physical requirement commonly called life.
I’m okay with the value of the dollar drastically decreasing. In fact, I think it will naturally do so due to the high debt and labor becoming more expensive in other countries.
I think it would be good for U.S. workers as it will help make them more competitive in a global labor market.
ANY call to set a debt ceiling needs to be directly tied to converting all for-profit corporations into public-benefit corporations.
That doesn't restrict corporations from generating profit. It just balances how they use said profits.
"Fixing" the national debt problem by putting a ceiling on it without addressing one of the main contributors to the current state of government spending and quality of life in America is a recipe for disaster.
It's okay to rip the band-aid off, but only if you're ready to deal with what may be a mortal wound.
Look at all of the major government spending and determine who benefits most and where funds are sent. If the recipient of funds is a corporation, also review how their profits are spent and which humans eventually profit from that.
1. As things get more technologically advanced/complex, the number of people who can tell you that a bad idea is bad keeps diminishing. Crypto and blockchain are classic examples. Financial experts told you crypto is going nowhere. Database experts told you blockchain is a nothingburger. But they were drowned out by the sea of laypeople (or experts in other things) telling you how crypto and blockchain would become the bedrock of tomorrow's society.
2. It is now common knowledge that one can become adequately rich from working on or investing in bad ideas, to the point where an idea isn't even considered bad if you can grift investor money from it. In other terms, it's no longer taboo to scam people richer or stupider than you; it's just "hustling".
I will wait.