Shree R

Innovate & Accelerate, ASI is coming | AI x Product Nerd | Techno-optimist 🚀

[Shree R @Shreezus42]

The key to innovation is extending opportunity to as many humans as possible. 🌍

[Sholto Douglas @_sholtodouglas]

If deployed widely - I bet this will save the US healthcare system at least 100x all of MJ’s profit to date.

It’s a great example of how much better someone like David is able to allocate capital than ~the rest of the world.

The AI boom should empower a generation of people who understand just how fast we can climb the tech tree, and will dream very, very big. Expect incredible things.

When Midjourney Dyson Spheres?

[Satya Nadella @satyanadella]

https://t.co/vLmiBKTtX3

[Awni Hannun @awnihannun]

It's very cool that Apple shipped a 20B parameter on-device.

You can't put 20B parameters in RAM at any reasonable precision. To make it work they are using pretty exotic architecture by today's standards.

A small model predicts from the query (or prompt) which experts to load from Nand into RAM. The key distinction from a typical MoE is that you do this once per query and then generate all the tokens with the same experts (instead of switching the experts for every token).

[Shree R @Shreezus42]

@sethrosen https://t.co/V9rTZQDhp6

[Blair Dulder CPA™ 🧃 @runaway_vol]

Nobody has ever made it to Series Z. If you can raise through Z with no down round, a16z will implode into a black hole and swallow the Bay Area.

[Bryan Johnson @bryan_johnson]

never waste a good crisis

[Shree R @Shreezus42]

@mustafasuleyman Awesome work from the MAI team 🚀

[Elad Gil @eladgil]

The events of the last 6 months in technology are arguable amongst the most important in human history

The tools now increasingly exist for recursive self improvement of models & agents

We are likely in very early lift off & exponential

Largely unnoticed outside of tech

[Shree R @Shreezus42]

@sporadica Human lifespans are going to extend far longer than most expect.

[Tanay Padhi @tanaypadhi]

how did Allbirds pivot to AI compute hardware before the shoe company literally called ASICS

[Shree R @Shreezus42]

@MorePerfectUS Yeah, let’s ban automated elevators to protect lift operator jobs while we’re at it. 🤡

Decels are morons

[Anthropic @AnthropicAI]

Introducing Project Glasswing: an urgent initiative to help secure the world’s most critical software.

It’s powered by our newest frontier model, Claude Mythos Preview, which can find software vulnerabilities better than all but the most skilled humans.
https://t.co/NQ7IfEtYk7

[Shree R @Shreezus42]

@johncoogan @sama You guys are gonna need a bigger gong

[Shree R @Shreezus42]

@mattparlmer You can just make things
🇺🇸

[Shree R @Shreezus42]

@tszzl Not bad!

[Shree R @Shreezus42]

@karpathy Similar to human memory, what an AI forgets might be just as important as what it retains.

[Shree R @Shreezus42]

@simonw @SumitGup You’re gonna have to run it overnight just to generate a haiku

[Satya Nadella @satyanadella]

We’ve trained a multimodal AI model to turn routine pathology slides into spatial proteomics, with the potential to reduce time and cost while expanding access to cancer care.

[Dwarkesh Patel @dwarkesh_sp]

The AI supply chain has the craziest value cascade of any industry in the world.

thinks that over the next five years, the biggest bottleneck to deploying AI will be EUV machines.

ASML sells EUV machines for $300-400 million. You need about three and a half machines, so $1.2 billion, to produce a gigawatt of compute each year.

So $1.2 billion of ASL and ML machines are producing $35 billion worth of NVIDIA chips a year (and of course those machines can be used over many years).

Then those Nvidia chips can generate $100s of billion of token revenue for AI labs.

Or think about: over the last three years, TSMC spent $100 billion on CapEx.

A small fraction of that went to building the fab capacity used for Nvidia's chips. Nvidia turned that fraction into $216 billion of revenue last fiscal year alone. Relative to the value of the thing they’re producing, TSMC’s CapEx looks crazy small.

So why aren’t ASML/TSMC/etc juicing up margins to capture more of the value they’re generating? And why aren’t they scaling up CapEx so that they can take advantage of the crazy AI growth worlds?

@dylan522p says that it’s because these upstream players aren’t AGI-pilled enough. They’re still looking for stable, gentle long-term growth. The insane demand signal is taking a long time to percolate up the supply chain.

[Anthropic @AnthropicAI]

Introducing The Anthropic Institute, a new effort to advance the public conversation about powerful AI.

https://t.co/M7vi9oRuYi

[Y Combinator @ycombinator]

Congrats to @maxhodak_ and the @ScienceCorp_ team on their $230M Series C!

Their PRIMA implant is letting blind people see again: improved vision in 26 of 32 trial patients with advanced macular degeneration. They're on track to be the first BCI company to bring a vision restoration product to market.

https://t.co/mX0cdpt5LV