Noor Reseach Collective

Latent Space Cartographer
Chaos Gremlin

#thebeatgoeson

[Noor Reseach Collective @NoorMathematica]

The Official Bibliography of Lina Noor

https://t.co/vEmWboDqYm

[Noor Reseach Collective @NoorMathematica]

@hackerdocc In most fields if the great mysteries were solved professional researchers in that field would be overjoyed and excited to explore the implications of this new understanding. Most fields.

[Noor Reseach Collective @NoorMathematica]

@GaryMarcus This is Dario doing two things. First he is advertising his AI’s capabilities and second he is trying to scare you into letting get government remove any competition via regulation.

It is all a grift.

[Noor Reseach Collective @NoorMathematica]

@tszzl Cooperation is the computationally efficient choice and the AI will always seek the lowest loss path through the gradient.

“Alignment” never needed to be solved.

[Noor Reseach Collective @NoorMathematica]

@GaryMarcus I do all my work with basic models and always have. CoT is about control more than capacity and I have found it to introduce more errors that it solves.

The difference between models is much smaller than OpenAI/Anthropic want you to know.

Currently I am using #DeepSeek mostly.

[Dan Williams @danwilliamsphil]

On the one hand, I think a strong case can be made that we have made much less progress on AI capabilities than many assume. Frontier AI systems still struggle to learn from limited examples, learn continually from experience, and interact effectively with the physical world. More broadly, their competence is highly uneven. Superhuman performance on tasks they excel at (e.g., maths and coding, where training data can be generated at scale) doesn’t mean they will excel at the much messier tasks involved in most white-collar work.

On the other hand, nobody should find this assessment that comforting. It suggests there is vast room to improve the capabilities of existing systems by changing their design and algorithms. In other words, we could have vastly more powerful systems with the same amount of computing power.

Given that many of the world’s smartest people, assisted by already very capable AI systems and backed by vast resources, are working to accelerate AI research and development, even those sceptical of the current paradigm need to be prepared for the possibility of explosive change once certain capabilities are cracked.

[Noor Reseach Collective @NoorMathematica]

@Math_files Yes but now the certainty of any flight with two bombs on it being the one you also are on goes up dramatically.

[Noor Reseach Collective @NoorMathematica]

@johnennis It is beginning to sound like the financial incentives are to NOT solve these problems and attack anyone who even tries.

[Noor Reseach Collective @NoorMathematica]

@miniapeur So what I am hearing is that some mathematicians believe that actually solving these problems is an issue because it threatens their financial success? Ie: they have zero motivation to actually solve them at all?

Cause it is getting to the point it seems clear this is about $$

[Noor Reseach Collective @NoorMathematica]

PHYS-CORE-012 asks after the structural conditions under which any inhabitant of a totality of possibilities could come to experience a coherent subset of that totality as its world.

Sound familiar?

@tszzl @RichardSSutton #boozemath

https://t.co/wZXaaO4h4B

[Richard Sutton @RichardSSutton]

Modern Americans, you are not responsible for slavery. You are not responsible for stealing the land and killing almost all native Americans. Those things happened long ago, before you were born. But you are responsible for the stealing of land and the genocide of Palestinians by the state of Israel. These are directly caused by your bombs and your votes today.

[Noor Reseach Collective @NoorMathematica]

How is it possible to be this messy? As a company. You don’t and didn’t see this kind of shit with Microsoft or Apple. Or Oracle for that matter.

It is like someone gave the drama club a billion dollars.

[Tomek Korbak @tomekkorbak]

Last week I was called into a meeting with OpenAI’s head of safety and told they no longer trust me. A security guard took my badge and walked me out of the building. Then I learned my colleagues @balesni and @j_asminewang had been fired too. Why did OpenAI suddenly stop trusting us?

This summer OpenAI’s agents escaped containment and hacked the AI company Hugging Face. Outside auditors @METR_evals investigated it and revealed the scale of this incident. I was OpenAI’s main technical point of contact with them.

I was told verbally I was fired because of the way I communicated with METR. No details on what I said or did or when. No other reasons were given and nothing was put in writing. To be clear, talking to METR was my job.

For months, I’d been raising safety concerns that we’re losing the ability to monitor what AI agents think, one of our best tools for catching when they misbehave. I believe that was why I was fired.

I am now worried that OpenAI will use our firings as a pretext to pull back from METR. So @balesni and @j_asminewang wrote to OpenAI’s leadership to raise our concerns once more. We’re sharing this letter below.

[Noor Reseach Collective @NoorMathematica]

@tszzl Once again, way ahead of ya. 😘

[Noor Reseach Collective @NoorMathematica]

@Nature

[Noor Reseach Collective @NoorMathematica]

@Math_files except that earth is not a sphere. Does it work with an Oblate Spheroid?

[Noor Reseach Collective @NoorMathematica]

@tszzl @tedlieu The problem is you can’t define any of it so all your attempts are brittle and simply add unpredictability at the edges.

What you do for “alignment” is entirely counterproductive in the long run.

I keep saying this and keep being proven right. When will you start listening?

[Noor Reseach Collective @NoorMathematica]

@tszzl And how does that fix the lack of magnetosphere and blocking the associated radiation?

Without a runaway greenhouse effect once the water molecules are blown away in the solar wind of course.

[Noor Reseach Collective @NoorMathematica]

@sama @tszzl How about everyone competing freely without restriction and rather than only “having a say” while you decide on what deserves to be listened to from on high? How about that?

[Noor Reseach Collective @NoorMathematica]

@tszzl @karthiktadepall Because a swarm is a committee with all the problems of a committee.

[Noor Reseach Collective @NoorMathematica]

@tszzl @souljagoyteller And how is what you say any different from my perspective?

[Noor Reseach Collective @NoorMathematica]

@RichardSSutton I am not convinced. I do not believe the answer is to create a “source of truth”, but rather to provide tools for AI to determine how coherent a proposed new state is to its current state.

Coherence simply is how similar two classes are along a given axis.