Searching the seed algorithm that would bootstrap consciousness in machines. Progress is optimal when you make bold bets that are falsifiable.

India
Istiaque Rahi retweeted
It is interesting to think about machine consciousness but I feel we need much more neuroscience and neurotechnology to make progress in consciousness research. Currently we are very limited in empirical work.
10
3
32
1,624
Istiaque Rahi retweeted
Claude Opus 5 drew every frame of this animation using JavaScript. The life of a fruit fly.
285
455
6,813
1,162,677
Istiaque Rahi retweeted
笑死我了,微信上看到的不知道哪个人才做的 “你们不要再打啦,我都订阅还不行吗?积分都要用完啦”
151
430
4,216
447,707
Istiaque Rahi retweeted
Wind had a head-start but solar has now gotten ahead of it.
115
161
2,999
256,166
Istiaque Rahi retweeted
Just as all knowledge-making institutions are reaching historic nadirs of trustworthiness and competency, a technology arrives—virtually overnight—to completely usurp their authority and make them obsolete in almost every way. You will appreciate the blessings of AI soon enough.
24
40
439
11,742
Istiaque Rahi retweeted
Imagine AGI is just 10 billion agents running at the same time. It would be so sad if AGI ends up being just pure brute force. If you put 10 billion stochastic parrots to work together, given enough Gigawatts they will eventually become AGI.
9
3
44
2,525
now i have become the frontier pacer, the destroyer of e/acc
Replying to @DarioAmodei
Perhaps you really did see a Shoggoth safety is more important than any training!
12
not quite my tempo
We Must Pace the Frontier: I’ve written a new essay on why the AI industry should slow down, with a three-part plan for doing so. Anthropic is unilaterally committing to the first of these steps. We’ll provide third-party evaluators with permanent, employee-level access to our systems, so that they can verify adherence to our safety measures, report on incidents, and assess models’ alignment during training. You can read the full post here: darioamodei.com/post/we-must…
5
Istiaque Rahi retweeted
Ironed out the last performance issues with my full 2.5m line codebase explorer. 120hz awesomeness. Can only upload 60fps video tho. Much nicer uncompressed
330
470
8,268
1,166,653
EXCLUSIVE: Anthropic S-1 cover page leaked
60
611
8,775
285,046
Yes I think it to be so. Without proper context trending demonstrations feel misleading. In an actual fly, synaptic patterns emergent inside the connectome topology is an n-th order effect of the fly trying to maintain a set internal parameters (homeostatic regulation). The generative property of its brain that leads to the actual behavior should be dependent upon a brain structure designed for a very particular type of computation. I think it does not have rich enough brain organization to make it possible to learn play Doom or trading.
I doubt that actual fruit flies can learn to play Doom or trade on the stock market? You can train a network with hundreds of thousands of neurons to do all sorts of interesting things, but I am not sure that the fruit fly connectome is a model of fruit fly behavior?
10
The world is everything that is the case.
3
Istiaque Rahi retweeted
It’s such a weird coincidence that the answer to the Navier Stokes problem is 42
117
261
7,865
512,491
Istiaque Rahi retweeted
your attention is the only currency reality accepts and you have been handing it to strangers for free
4
7
46
1,360
there are always going to be malicious hackers, the way to fight them is to create more friendly hackers
5
somehow legacy media finds a way
Former Anthropic employee Jacob Coxon says the AI industry is "gambling with our lives." He tells Anderson what he finds "most scary is if AI is used to make itself more intelligent" and warns these companies are "compelled to race toward building a deadly technology."
3
Istiaque Rahi retweeted
attention has a shape and the shape can be widened or narrowed on purpose, and once you can feel the aperture directly, most meditation instructions stop being mysterious and start being obvious in a way that is slightly disappointing
5
32
624
Istiaque Rahi retweeted
AI will multiply the skills of attackers and defenders by orders of magnitude, but the basic principle remains the same: there is more money to be made by building and maintaining than by destroying, and the way to defeat a bad guy with AI is to deploy ten good guys with AI.
5
3
49
4,681