ilaksh

34 ms·

481.
by ilaksh 1y ago
Opus (and Sonnet) 4 obviously came out before they started the experiment.
482.
by ilaksh 1y ago
I've always felt that the ideal would be to somehow create many different tokenizations for different use cases. And allow them to sometimes build on each other. Like a lot of domain-specific languages. Maybe there could be something l
483.
by ilaksh 1y ago
What kind of models do they use for reviewing the actual plans visually? A normal SOTA VLM like Gemini 2.5 Pro? Or do they stop short of that and just make sure that all the required documents are included in the submission? I suppose for b
484.
by ilaksh 1y ago
My views are not naive. Let's get specnific. What do your propose? That LLMs or image generators be banned? Are technologists creating AI swarms for political manipulation? Or is that being done by politicians or political groups? Are
485.
by ilaksh 1y ago
Guns, drugs and religion are not equivalent to AI. I did not say there should not be regulation or standards. The point is that the whole diagnosis is wrong. People point at AI as creating a new problem as if everything was okay. But everyt
486.
by ilaksh 1y ago
Stop blaming technology for the way humans misuse it. AI, like any technology, is a lever. Like a big metal rod. You could use that to move stones for building a structure, or to dislodge a boulder to roll down a hill and destroy someone el
487.
by ilaksh 1y ago
It makes me question your honesty. If you want a fox logo then build it as a Firefox fork. If you do that I will trust you again.
488.
by ilaksh 1y ago
Yeah. Regardless, it seems misleading to use that icon with a Chromium fork. Also the fact that it's AGPL means this project is very copyleft and not compatible with business models. I'm not saying that there is no place for copyl
489.
by ilaksh 1y ago
Reminds me of VB6, which was amazing.
490.
by ilaksh 1y ago
Obviously there are going to be narrow tasks where fine tuning makes sense. But using leading models for agents is a completely different mindset and approach. Because I have been working on replacing multiple humans handling complex busine
491.
by ilaksh 1y ago
Maybe because they also are releasing o3-pro.
492.
by ilaksh 1y ago
My interpretation of this is he feels that AI-assisted programming is okay if you still participate in the programming sometimes, but if you only prompt then you're not really a programmer. Or something. He does say that you should pro
493.
by ilaksh 1y ago
There is urgent need for education on these types of topics in the United States.
494.
by ilaksh 1y ago
Interesting idea. Aren't there a few models a little more like what he suggests than a typical LLM? Like one or two experiments that operate on raw bytes, or some robotics diffusion transformers or whatever like Nvidia's thing? I
495.
by ilaksh 1y ago
I think those CEOs like Altman and Amodei do find it convenient to hype their products like that, but also they believe in computing exponentials and artificial superintelligence etc.
496.
by ilaksh 1y ago
The biggest reason there is such a difference of opinion on this is that people have fundamentally different worldviews. If you have bought into the singularity concept and exponential acceleration of computing performance, then you are lik
497.
by ilaksh 1y ago
Exactly how long did it take you? And now how much actual time was spent in the comparison prompting and code review by Cloudflare?
498.
by ilaksh 1y ago
It actually comes down to feedback loops which means iterating on software being used or attempting to be used by the customer. Chat UIs are an excellent customer feedback loop. Agents develop new iterations very quickly. LLMs can absolutel
499.
by ilaksh 1y ago
Anyone have a guess as to when the open dev version gets released? More like a week or a month or two I wonder.
500.
by ilaksh 1y ago
Look more closely at the example. Clearly there is an opportunity for inference with objects that only partially obscure.
501.
by ilaksh 1y ago
I think they did make an announcement on WeChat.
502.
by ilaksh 1y ago
https://github.com/runvnc/mindroot . Basically open source Custom GPTs. Heavily focused on plugins. Has browser use, computer use, agents can delegate subtasks, Chroma KB, etc. Fully customizable front end.
503.
by ilaksh 1y ago
How does this sort of thing perform with 70b models?
504.
by ilaksh 1y ago
He asked that in the context of an actual specific project. It did not make sense way he asked it. And it's the executive's to plan that out five years down the line.. although I guarantee you none of them are trying to predict th
505.
by ilaksh 1y ago
You might want to study the history of technology and how rapidly compute efficiency has increased as well as how quickly the models are improving. In this context, assuming that humans will still be able to do high level planning anywhere
506.
by ilaksh 1y ago
That's a completely nonsensical question given how quickly things are evolving. No one has a five year project timeline.
507.
by ilaksh 1y ago
What model does it use? gpt-4.1? Or can it use o3 sometimes? Or the new Codex model?
508.
by ilaksh 1y ago
This is a direct threat to Cursor. The smarter the models get, the less often programmers really need to dig into an IDE, even one with AI in it. Give it a couple of years and there will be a lot of projects that were done just by assigning
509.
by ilaksh 1y ago
I think they mentioned it was a similar environment to what it trains on, so maybe they have a default Dockerfile. Of course containers can also install additional packages or at least python packages.
510.
by ilaksh 1y ago
I don't think jobs are necessarily a good plan at all anymore. Figure out how to leverage AIs and robots as cheap labor, and sell services or products. But if someone is trying to get a job, I get the impression that networking helps m

More ›