brettkromkamp

So it begins.

What’s new? I mean, it’s the same bunch of companies propping up OpenAI because they know that if OpenAI fails this whole AI house-of-cards will come tumbling down around them.

Is there any (major) AI-company that is not a total douchebag? The greed on display combined with a total absence of any kind of moral compass is sickening.

Can the UN focus on something else than Israel?

Just trying to figure out why anyone would want to go to the US, as a tourist, given the present situation.

This would have to lead to model-collapse at some point, right?

Kicked out? Why wasn't he sent to prison? This is basically spying, right? I can imagine that if you do this kind of stuff in Russia you would rot in a prison cell (and, perhaps even worse).

It's so obvious that it's cringe. They have a product to sell so these kind of things are all part of a larger marketing campaign in preparation for their IPO.

Not only that... I would really have to see something tangible from a company that makes these kind of claims. ASML's technology is some of the most complex and advanced in the world and has it has taken decades to get there. Of course it is possible, but without something tangible to back up their claims, I think I will remain skeptical.

I so recommend reading this article. It will provide you insight into a world that, I imagine, the majority of readers are not familiar with: Python development in the financial/banking system.

This is definitely not the first time this article has appeared on HN's frontpage, but every time it does, I re-read the article because I find the approach to software that the author describes in the article to be fascinating. YMMV.

What am I missing here? How does this get to HN's front page?

Where’s the news in this? Would he say anything else than it not being a bubble?

It's a mixed bag. It depends on your problem domain, the problem you are trying to solve (within that domain), the context you provide the LLM, the output it generates (you are using libraries to coerce the output into predictable (JSON) structures, right?). What's more, based on what you are trying to do, the LLM you are using might have sufficient training data, but not necessarily so (resulting in possible hallucinations/confabulations). So, there's that. Also, LLMs are not deterministic, they can (and will) generate a different response every time you call them (even if the context you provided is the same). So, yeah... sometimes these things really deliver and other times, it's just... meh!