dei

Advancing global intelligence. Opinions are my own.
Working on the architecture that comes after transformers / interested in quantum photonics & emerging tech

[dei @dei1022]

Introducing a new mini-benchmark: Anti-Template Reasoning Test (ATR-T) v1.0: A proof-of-concept textual benchmark testing literal attention, template resistance, and abstention discipline on trivial items.

[dei @dei1022]

@AvmenLondon @zerohedge You know Meta has some of the largest compute clusters and datacentres in the world right…?

[dei @dei1022]

@davidwroe Let's hope our gov pushes for this, surely our gov is not so short sighted as to let this slip.

[dei @dei1022]

Don’t be fooled and drop your guard, this is last minute face saving from the US gov to prevent mass AI divestment from the USA. They showed their eventual ultimate hand with AGI/ASI far too early, thinking everything is okay now long term is a mistake.

[Anthropic @AnthropicAI]

We’ve received notice that the Department of Commerce has lifted export controls on Claude Fable 5 and Mythos 5.

We'll begin restoring access tomorrow, and will share an update soon.

We’re grateful to our users for their patience, and to everyone who worked with us on redeploying the models.

[dei @dei1022]

@DaveShapi Well said

[dei @dei1022]

@hopes_revenge 😔

[dei @dei1022]

@VladTheInflator Ironically, proposing 120 as the floor is kind of low IQ in itself. Even 110 would have a profound long-term net positive effect on America’s average IQ.

[dei @dei1022]

This point is very important to me and the work I have been doing behind the scenes. I hope to have updates that I can bring within the next 6 months, because there have been very promising results so far. I am happy to share that the lab is setup and operational.

[dei @dei1022]

The point of divergence has arrived far sooner than I expected.

Open source is more important than ever, but maybe “open source” isn't the true savior.

Models will be released that fall under custom licenses, that act as an equalizer for humanity, preventing elite capture.

[dei @dei1022]

@rob_mcrobberson If you say enough buzzwords surely that counts as shipping a product right?

[roon @tszzl]

@chris_j_paxton bearish only insofar as we are shooting long past AGI and much crazier things are coming

[dei @dei1022]

@basedjensen The problem that arises is how the governments will respond. That’s when we start getting real dystopian with compute restrictions, model bans etc.

We must hope the democratisation is rapid enough to be irreversible.

[dei @dei1022]

I wonder if this charter will be invoked in the next 18 months, or if OpenAI will reveal their hand before then?

[dei @dei1022]

Something interesting shipping from China soon..

[dei @dei1022]

Is the new AI psychosis people thinking every post is written with AI?

[dei @dei1022]

@tszzl AGI was the friends we made along the way

[dei @dei1022]

Some skepticism like this is good though as it helps ensure that the most important technology in human history will be done right. When billions of dollars are involved, its not really unsuited as that sort of money creates bad incentives (like benchmaxxing) for investor appeasement as “number go up = good” - especially with potential IPO’s.

Plus, a lot of the biggest strategic skeptics are also some of the biggest optimists for AI’s potential, hence being harsher with skepticism to help move it in the right direction.

[dei @dei1022]

I don’t see the innate appeal of transhumanism, especially it brings up uncomfortable questions.

If you upload your mind, is it really you? Or does this instance of you die and a copy lives on? And some may say that you will not be able to tell the difference because you die and wont wake up to perceive it, but no sane human would willingly sign up to die so that a copy of them gets to live on a little longer.

If you start replacing your body, how many parts until you are no longer you? Do you reach a point where you just die after something vital is accidentally replaced? Does our specific consciousness rely on the specific conditions in our brain, and perhaps even the unknown quantum interactions that occur? Do you become the new ship of Theseus problem?

And I agree, whats the point of inserting the low tech meat brain into the middle of the machinery? If ASI is truly ASI, even if we can think at the same speed as it with Neuralink etc, we are only going to be a bottleneck. Is the current paradigm of helpful assistant not the ideal, where the AI is unhindered, and your humanity is still yours?

[dei @dei1022]

The era of compute restrictions is looming. Are you hiding compute under your floorboards?

[dei @dei1022]

Bread and circus is in full force… not a good sign right now

[roon @tszzl]

1. if transacting with superintelligent models outside of the boundaries of a lab becomes difficult due to national security / ai safety concerns and so on, it will mean the Coasean boundaries of the labs will grow to encompass all interesting industry, creating a truly cyberpunk chaebol-capitalism type of future, where the goverment sort of runs them but they also sort of run the government

2. as if there weren't already enough reasons to break up your family, leave your home, the Zone of Thought will increase the attractiveness of migrating to try and have your child on american soil, so they can have 1000x the effective brain power of people born elsewhere

3. every country should probably try and either work towards a new ai security pact with the americans immediately or pool every ounce of national resources to try and create their own ASI labs lest you become complete intellectual, economic, and moral vassals to the united states of america and the output byproducts its ASIs (you wont even get to talk to them). if they succeeded (big if) this will imply a more global race and more risk factors than was previously implied by the formerly only "beating china" narrative -- but many will prefer it to the superintelligent monopolar value lock-in

4. the other alternative is to keep the tension between safety and concentration of power at the top of mind and for the government/labs to push for solving it, rather than instrumentalizing all other values to be subservient to minimizing ai harms. insofar as safety means defending properties of the fragile world we like, the diffuse nature of power is one of those properties

5. historically the americans have been really quite Benign about their global public goods hegemony despite the ability to extract significantly more rents than they do, and it makes it easy for people of all stripes to fight for america rather than under it. we probably don't have to, but i hope america overall works towards export promotion of american models rather than export control