avaer

23 ms·

511.
by avaer 9mo ago
The difference is LLMs let you "run Google" on your own data with copy paste. Which you could not do before. If you're using ChatGPT like you use Google then I agree with you. But IMO comparing ChatGPT to Google means you hav
512.
by avaer 9mo ago
Don't forget porn. Though it's a use case people like Satya will want to avoid for reasons.
513.
by avaer 9mo ago
The PWA is a nice touch.
514.
by avaer 9mo ago
If I had some heartfelt advice for the Mastra devrel team it would be to shut up about about Gatsby. I'm a happy Mastra user and I'm biased to their success. But I think linking it to an unrelated project is only going to matter t
515.
by avaer 9mo ago
It's definitely cleaner than what you will see in the real world. Research-quality repositories written in partial Chinese with key dependencies missing are common. IMO the assignment('s purpose) could be improved by making the co
516.
by avaer 9mo ago
Kind of, but not any particular GPU. The machine is fake and simulated: https://github.com/anthropics/original_performance_takehome/... But presumably similar principles apply.
517.
by avaer 9mo ago
It's pretty interesting how close this assignment looks to demoscene [1] golf [2]. [1] https://en.wikipedia.org/wiki/Demoscene [2] https://en.wikipedia.org/wiki/Code_golf It even uses Chrome
518.
by avaer 10mo ago
What's the difference between AI ads and computer-assisted bribery? We can wrangle the legalese (as AI companies certainly will) but is there any ethical, moral, or practical difference?
519.
by avaer 10mo ago
To be fair, even if "from scratch" means "download and build Chromium", that's still nontrivial to accomplish. And with how complicated a modern browser is, you can get into Ship of Theseus philosophy pretty fast. I
520.
by avaer 10mo ago
I'm don't blame you for the mistake, but if AI can't write a correct HN comment maybe execution (ie code review, testing, reading what you're posting) matters more than you think.
521.
by avaer 10mo ago
Seems cool, but is there more to this than just some outlines in a graph? That's all I was able to try without paying. You mention there is "fountain exports" so I was expecting something more akin to a live-morphing screenpl
522.
by avaer 10mo ago
No, Github should pay open source per inference token. It'll never happen; open source doesn't have the legal team of Disney [1]. [1] https://openai.com/index/disney-sora-agreement/
523.
by avaer 10mo ago
The comment wasn't referencing LLMs, but generative AI. Even then, given the deep impact of LLMs and how many people are using them already, it's a stretch to say LLMs will have no effect on the development of AGI. I think it&#x27
524.
by avaer 10mo ago
If AGI is ever going to happen, then it's definitionally a precursor to it. So I'm not really sure how to parse your statement.
525.
by avaer 10mo ago
Pretty sure they knew what they wrought.
526.
by avaer 10mo ago
I wish governments would treat things like this as equivalent to fraud. It might not be taking people's money, but it's taking away access to truth through automated deception, which might even be worse. Back when it took a madman
527.
by avaer 10mo ago
> Anything that removes the power of CEOs and gives it to the worker should be highly encouraged. Except replacing CEOs with AIs will not do this.
528.
by avaer 10mo ago
Last year's joke is next year's reality. Every year I feel a bit less crazy in my silly armchair speculation that the Second Renaissance from the Animatrix is a good documentary. If AI "takes over" it will be via economi
529.
by avaer 10mo ago
Ah, but if we're not talking 6DOF what's new with ml-sharp? We've had good autostereoscopy for a couple of years at least. > So far the best looking results are still achieved with good old mesh warping and no inpainting a
530.
by avaer 10mo ago
This is a free research project on GitHub. I think I'd rather apple focus on making hardware than hoarding GPUs for PR stunts to prove they are a "serious company".
531.
by avaer 10mo ago
It doesn't but it's pretty trivial to do if all you want is a pinholed mesh. I managed to one-shot it by mixing in the mesh exporter from https://github.com/Tencent-Hunyuan/HunyuanWorld-Mirror but at that poi
532.
by avaer 10mo ago
Unfortunately not as significant as you'd think. The output is not automatically metrically scaled (though you can use postprocessing to fix this, it's not part of this model). And you can't really move around much without ge
533.
by avaer 10mo ago
Efficient realtime video diffusion will revolutionize the way people use computers even more so than LLMs. I actually think we are already there with quality, but nobody is going to wait 10 minutes to do a task with video that takes 2 secon
534.
by avaer 10mo ago
I would never talk down on Rob Pike. But I think in the aggregate ChatGPT has solved more problems, and created more things, than Rob Pike (the man) did -- and also created more problems, with a significantly worse ratio for sure, but the p
535.
by avaer 10mo ago
Maybe not autonomously (that would be very close to economic AGI). But I don't think the big companies are lying about how much of their code is being written by AI. I think back of the napkin math will show the economic value of the o
536.
by avaer 10mo ago
I used to write web apps in C++, so I totally understand not caring if it gets the job done. I guess the difference where I draw the line is that LLMs are inherently random I/O so you have to treat them like UI, or the network, where y
537.
by avaer 10mo ago
> On the other hand I am trying hard to remember anything famous created by any LLM. That's because the credit is taken by the person running the AI, and every problem is blamed on the AI. LLMs don't have rights.
538.
by avaer 10mo ago
It's still not true for any complicated extraction. I don't think I've ever shipped a successful solution to anything serious that relied on freeform schema say-and-pray with retries.
539.
by avaer 10mo ago
To this day many good models don't support structured outputs (say Opus 4.5) so it's not a panacea you can count on in production. The bigger problem is that LangChain/Python is the least set up to take advantage of strong sc
540.
by avaer 10mo ago
I somewhat take issue as a LangChain hater + Mastra lover with 20+ years of coding experience and coding awards to my name (which I don't care about, I only mention it for context). Langchain is `left-pad` -- a big waste of your time,

More ›