typon

When you call yourself "Open"AI and then turn around and backstab the entire open community, its pretty hard to recover from that.

No this is well known. Look for Table 2.2 in GPT3 paper.

Its a shame this model is performing so well because I can't in good conscience pay money to Elon Musk. Will just have to wait for the other labs to do their thing.

LLMs are all built on stolen data. There is no such thing as intellectual property in LLMs.

They have a large robotics research group.

thanks for the recs!

I wish I could experience the feeling of reading The Count of Monte Cristo again

That's religion, not a predictive theory.

The Bitter Lesson has held up in a lot of domains where injecting human inductive bias was detrimental. Adding LIDAR for example is not inductive bias - it's a strictly superior form of sensing. You won't call a wolf's sense of smell "hand engineered features" or a cat's reflexes a failure of evolution to extract more signal from an inferior sensory input.

Waymo will win because they want to make a product that works and not be ideological about it - that's ultimately what matters.

The counter argument is a bitter lesson that Tesla is learning from Waymo and the lesson might be bitter enough to tank the company. Waymo's approach to self driving isn't end to end - they have classical control combined with tons of deep learning, creating a final product that actually works in the real world, meanwhile the purely data driving approach from Tesla has failed to deliver a working product.

Gotta love these guys for keeping at it for 26 years. Much much love

Imagine the actual reverse: police surveil American jews for attending pro Israel/pro genocide protests. The police chief would be fired within a day.

Creating an underclass that relies on economic elites paying taxes rather than being economically independent because you want to optimize for "high value add industries" is a terrible long term strategy.

Imagine if the world's engineering talent was focused on this rather than making AI to generate slop?

Well no because Leela today beats human + AI combos or humans only just as well.

With AGI/ASI, there will come a point where this paradox breaks down: once the ASI is self-sufficient and "agentic enough", people's direct use for it will actually reduce. You won't need to "micromanage" it - and the total amount of compute it requires might actually drop drastically if the ASI improves its own efficiency much better than humans can. At that point the question becomes one of alignment: will overall consumption drop because ASI subjugates or eliminates humanity or does it grow exponentially as we use it to explore the stars.

OpenAI is the MIC darling - expect more ridiculous attacks on competitors in the future

Altman rising to the top and becoming the defacto state preferred leader of AI in the US is wild. Fair play to him.

Historical revisionism is supposed to happen when enough time has passed.

This ceasefire is directly due to Trump's pressure on Netanyahu. Trump's envoy went to Doha and made this happen: https://www.axios.com/2025/01/11/trumps-israel-gaza-deal-jan...

This is the actual reality

Just replying to the above comment that seems to suggest that all these DEI jobs are being taken over by "black or Hispanic" people.

In my entire career working for US companies, I have yet to work with a black software engineer. Not a auxiliary role like PM, DevOps, IT but a straight SDE role. I have worked with literally hundreds of software engineers in my life.

This is one technology that I've been waiting for for over a decade. Apple apparently hasn't given up on it, despite working on it for 10+ years as well. I am hopeful it can be solved because diabetes is one of the biggest killers on Earth today.

The rules are written by the powerful to advantage the powerful. Once those advantages disappear, the rules must be rewritten or disregarded.

Xenophobia

Maybe you don't understand how this works?

It's able to query the relevant documentation, put it in its context and then use that to generate code. It's extremely relevant to giving existing models superior functionality.

Asked it to generate code for a library that was released in the past year - GPT-4 couldn't do it and this one just did it flawlessly. I am super impressed.

Weird that this article doesn't mention Carrom at all - maybe people in North America haven't heard of it.

If her contrarian view is correct, why shouldn't she be cynical since no one else seems to care. If her contrarian view is incorrect, then sure don't listen to her.

Does he have some damning videos on his investors? I am truly baffled as to why this guy keeps getting funded

My modelling the relevant parts of the worm that are involved in these processes.

You don't need an atomistic DFT simulator to get a good enough simulation of fluid dynamics, enough to design jets or rockets that go into space.

I can imagine a world where you can simulate a simple organism like a worm without knowing how to simulate the cells that make it. This is because emergent behaviour of complex systems can be much simpler than the constituent parts of that system (temperature vs. knowing particle velocities).