Absolutely. I see the IA as more of a bulk storage area, with little community, and my site will have more commentary and community built around it. But everything that gets scanned for my site will also be available for the IA too.
I am still dreaming of something that would allow a user to file a ticket, have them record audio and video like loom to describe the issue and what they were trying to achieve, and then dump a screen record of the last minute before opening the ticket as well as as much info about the machine's state as possible. And/or maybe connecting to helpdesk with video directly. Existing software comes close but is not quite there yet.
Azure DevOps has a browser extension that can this except record audio of the person speaking what’s happening. Also, the user experience is fine for like… power users, but it’s not super fun to use.
The first satellite to fly W Band was in 2021. It supports much higher data rates but has atmospheric attenuation issues. This would be the first commercial use of it. It's possible SpaceX is using this to reduce downlink bottlenecks.
Excellent callout. Might be a component of Starshield. Hard to say until the RF is active in the wild or FCC filings are public (is it possible SpaceX is filing in Tonga to sidestep FCC public filings on the tech?). Would be wild if this system will be able to provide global airspace radar surveillance to the US DoD with high precision.
Do these implementations use the neural engine? I saw that there was a stable diffusion implementation using the neural engine and I found that my macbook noticably did not run hot, as opposed to an average Teams call.
It doesn't. You need to generate models for use on the neural engine, which apple did for Stable Diffusion, but this is just taking advantage of lots of fast RAM and lots and lots of threads, if I understand it correctly.
It uses Metal acceleration, and takes advantage of the shared memory architecture, meaning it's basically a GPU with 196GB VRAM. Trading space (VRAM) for time (FLOPs), it can beat the performance of an RTX4080 here.
Encoder only transformers (like BERT) can be made to run on neural engine with CoreML. Efficient inference with autoregressive encoder-decoder and decoder only transformers (aka LLMs) needs KV-caching, which currently can't be efficiently implemented with CoreML (and thus neural engine). So, for now it's GPU only, with Metal.
You can do autoregressive decoding with KV caching on the Neural Engine. You have to make a bit of a trade off and use fixed size inputs [1] but the speed up over no caching is meaningful.
There's a Whisper (Encoder-Decoder) [2] implementation if you want to see it in practice. Shameless plug, but I have a repo [3] where I'm working on autoregressive text generation on the Neural Engine. I'm running gpt2-xl (1.5B params) locally with KV caching at 120ms/token (vs. 450ms without caching). Will push an update soon.
Without quantization you can't go much higher than 1.5B params on M1's Neural Engine. M2 seems to have a higher ceiling but I haven't measured. I'm optimistic (but have not tried) that the new runtime quantization added to CoreML this year will allow for larger (and maybe faster) models on both.
Autoregressive transformer models are usually memory bound, whereas SD is compute bound, so perhaps the difference lies here. Also the reason why SD runs so much faster on the GPU than on the CPU.
M1 has (fast) unified memory between GPU and CPU, so something being memory bound ought not to have much bearing on whether it belongs on CPU or GPU… at least in theory. I’m a total noob here though so I may be wrong.
No. On the contrary, GDPR explicitly allows data to be stored if retention is required for law, such as all financial data for 10+ years. However, there is absolutely no need to hold all this data "hot" in the production system and not in a "cold" archive without automated connection.
Thats the whole point of the initial comment. It is necessary to store some potential PII for long periods of time due to regulations from the same people that want you to get rid of all PII.
What are the warning signs? The paths you described kind of describe things I do. Sometimes they are necessary and I am taking steps to prevent them. At the same time I worry I already shut down the warning signs.
Sometimes I wonder whether "burnout" or "occupational burnout" describes something I have experienced in college. It was not a spectacular or intense thing, just losing interest in my subject completely and avoiding getting started with exercises, even though I was decently good at my subject and loved the hours and hours I spent studying.
The usual, some combination of sleeping difficulties or sleeping too much, constantly being annoyed by nothing in particular, brain fog, forgetting things, procrastination, difficulty reaching a flow state, general disinterest in certain or most things, acute anxiety/stress response to thinking about things that should be done, increased "self-medication" with alcohol or other drugs, comfort eating, or alternatively forgetting to eat, loss of appetite.
This is me since... at least high school. Probably earlier, but my memories are spotty. Almost all the things you mention present, often acutely, in relation to anything that is "to be done" - be it work or personal. Anything can trigger it, the moment it stops being something I do on a whim, and becomes something that I plan or is expected of me. No underlying medical condition to attribute it to. What to do then?
Yes. Treated for depression in my early 20s, but it only reduced those issues a tiny bit. Diagnosed with ADHD in my 30s (thanks to many comments like yours over the years on HN, which eventually made me talk with a doctor again); treatment was much more effective for the symptoms in question, but they never completely went away, and occasionally come back to bite with vengeance. I currently consider the effect of that treatment to be somewhere between "close, but no cigar", "you're holding it wrong" and "lulled into false sense of security".
One thing I observe being unstated for many mental health conditions is that the duration is often life.
Or that without fundamentally changing the set and setting of one's life, one can expect, on average, approximately the same metrics now as tomorrow, and so on and so forth.
Therapy and medication help, for some people, some of the time.
> Sometimes I wonder whether "burnout" or "occupational burnout" describes something I have experienced in college. It was not a spectacular or intense thing, just losing interest in my subject completely and avoiding getting started with exercises, even though I was decently good at my subject and loved the hours and hours I spent studying.
I think this is my experience right now. I love CS and almost all my classes were great, but now it's like I can't study, I can't will myself to work on a class' project, etc. I have an exam that I have been avoiding for months and it's blocking my other classes that require this exam to take their exams
The fact that you use the past tense means you got past it? If so, I'd appreciate any advice
I felt the same with Electronic Engineering in my final year - because I had realised that what I was learning was so academically focused that it was almost useless in the real world. I forced myself to finish that last year, but I think that effort destroyed my love for electronic design work (fortunately, I fell into a software job instead, in part because I got my degree).
Perhaps if you can get out in the real world then you will find real problems and those will likely motivate you (if you are anything like me, anyway). The most motivated students I recall were already working, and they picked and chose relevant academic focus that could help them with their design work (i.e. they could get some value from the academic system). Even though work is often depressing in itself (varies on a huge number of factors).
Edit: perhaps relevant: I was depressed due to a relationship. She ended the relationship with me, and the next day I was long-term happy. Turns out situational depression is a thing, and that it is entirely different from clinical depression (which doesn’t fix itself in a day - exception fast bipolar?). If you feel unhappy, sometimes you have the ability to fix the situation that is making you unhappy!
Whenever I bring up the topic of Georgism or land value tax, people look at me like I am mentally ill. The idea of taxing the right to exclusive use over a piece of land - something that no human has ever put effort into existing - is so utterly foreign to most people. Unfortunate.
There was a reform of the German property tax system, since the federal constitutional court ruled that the old system was unfair. The current model allows the 16 states to set up their own system, if they wish so, and interestingly enough, Baden-Württemberg turned the property tax into an actual land value tax, taxing only the land value itself, not the real estate on it. Of course there was a documentary the other day about a land owner crying that he is forced to build more houses to rent out and cover the cost of the LVT. That was the point all along!
I disagree with it not being easy to understand, on the contrary it makes perfect sense to me. We have the vast earth we live on as humanity and all the land was here long before us and will be here long after we are gone. Now we divide some of this land up and give people exclusive rights to it. How is it not absolutely intuitive to pay everyone else, who is not allowed to use this land, a fee, rent or tax or however you might want to call it, as compensation? In turn, if you use less land to reach the same goal, i.e. you make better, more efficient use, you pay less, just like with any other resource.
The more common "rebuttal" I hear is "well the land lords will just slap it onto the rent, so no use", to which I ask "then why don't they ask for more rent already if the rentors would be willing to pay?"
Then again, people I talk with are against a wealth tax as well, claiming that property would be impossible to account for and value, meanwhile France and Switzerland, two neighbours, have no issues in that regard. At the same time, both Germany and Austria had a wealth tax until very recently.
I like your rebuttal, that is a very simple way to explain it. I’m not very good at explaining concepts so maybe my difficulties are with my method of explanation and not the subject.
