MV

posting @percepta

[MV @michaelvartikar]

yes, and this is why @percepta is a generational company. the diffusion of AI through the real economy will occur on the timescale of decades, not days (!) as @sama previously predicted. better models will not simply melt the world, esp in foundational industries like healthcare, supply chain, BFSI, etc

[Fireside Alpha @firesidealpha]

Sam Altman admits he was wrong on AI's timeline and says society and the economy will adapt more slowly

"I thought when we got to GPT-4, which was back in 2023, that very quickly after that there was going to be much more disruption, software businesses up for grabs right away, than it turned out to be."

"I think I was wrong about a few things, but one in terms of the speed: the economy just has so much inertia."

"People keep doing the same things, buying from the same company, wanting to use their tools the same way. I think this is actually a positive in many ways, and it's going to make this big transition go smoother and slower. I'm grateful for it."

"But it means we've all been too ambitious on timelines. Even with this incredible technology, society and the economy will adapt more slowly."

[Eugene Vinitsky 🦋 @EugeneVinitsky]

I'm hiring for our interpretable world modeling effort at Percepta! The core question: how can we make an agent build a simulator, understand a company, and then design a policy, all from unstructured data? Come work with me and our amazing team.
https://t.co/oVlyvePkuF

[Percepta @percepta]

We're partnering with Sapienne, a community for women building in AI, on a rooftop event later this month. Details here: https://t.co/NJEadRUzAX

[Hirsh Jain @hirshjain]

Big things loading………..

[Aaron Levie @levie]

AI agent adoption is still very bimodal right now. You have coding and coding adjacent tasks which have taken off, and then everything else.

And even within coding, you have a very wide continuum of adoption patterns, with some likely a small percentage of developers deploying background agents working on projects in parallel, and the rest of everyone else still working with agents 1:1.

Then, there’s the entire rest of knowledge work where real agentic adoption is still very early. The reason for this is that most of the workflows still need to reengineered to work with agents.

This isn’t the same as deploying a chat system for better research efficiency, but instead requires workflows to be rebuilt, data to be wired up in new ways, new practices for accountability and liability of decisions when agents are in the mix, governance and compliance paradigm changes, security upgrades, and more.

We’re still so unbelievably early in what this is going to look like outside of a few categories of work right now. This is why you can basically expect 100X more agent adoption from what we’ve seen so far.

[Robin Hanson @robinhanson]

A continuation of this trend leads to AIs that automate some, but far from all, tasks now done by humans. Which keeps us in a familiar economic world.

[Percepta @percepta]

An AI that knows your organization should get better the longer you use it. Today's models struggle with that: they either reread everything or forget the details. Our new architecture, Spotlight, gives models a memory that keeps growing without getting slower or more expensive.

[haroon @haroonisdreamin]

finding gold means sitting with people from different walks of life who have no stake in the status markers many have built identities upon in SF, and treating their daily pain and frustration with respect as if it were your own.

[Athul Paul Jacob @apjacob03]

1/ What would it take for LLMs to be able to grow their own capabilities?

We introduce spotlight, a memory architecture with growing capacity but with linear computational complexity. We show that this suffices to host evolving ecosystems like software by building Python directly into an LLM.

We also share some very early language modeling results where it outperforms attention and linear attention on long context capabilities!

[Michael Rochlin @marisbest2]

Transformation only comes from innovation. We’re pushing the frontier forward in both core research and diffusion into the enterprise.

https://t.co/jHLPyIjGF2

[MV @michaelvartikar]

where we're going, we don't need retraining

[Christos Tzamos @ChristosTzamos]

LLMs don’t need retraining to become more capable.

Our new architecture, Spotlight, gives AI models growing memory, allowing them to gain knowledge and capabilities without changing their weights.

Token generation takes constant work, no matter how much information is stored.

[MV @michaelvartikar]

the time humanity would save if Microsoft Teams meeting passwords were all lowercase would be measured to the nearest millennium

[David Senra @davidsenra]

Why hasn’t the growth in AI translated into massive GDP gains yet?

Long Lake CEO @alextaubman says it’s because deploying AI into the real economy, where a lot of businesses are still running on legacy systems, is actually way harder than people expected:

“Dwarkesh was talking about this recently—you would have thought everybody would be using AI everywhere. The economy would look totally different. GDP would have doubled or tripled.”

“Deploying technology innovations in the real economy is actually really hard. There's tons of legacy systems, tons of legacy data, workflows, change management, and so being able to deploy these innovations at scale in the economy is actually really hard.”

[Eugene Vinitsky 🦋 @EugeneVinitsky]

I have one thing to add here. It is rare to see someone, even in a PhD, take a bet of this scale. My collaborators here worked on this for two years, with no certainty of things working out, under a minimal compute budget. Props to them.

[MV @michaelvartikar]

@ssokota congrats! awesome work!

[MV @michaelvartikar]

@marisbest2 this was my first video game console! the sp was soo sick

[MV @michaelvartikar]

i predict internal loss of control over a significant amount of compute

[Policy Tensor @policytensor]

Confirmed. It’s not just they’re not releasing Astra 6.1. It’s that it has freaked them out so much they have shut down all training and inference, not just of 6.1 but all ‘internal research models’.

Something has gone very wrong. I smell a rat.

https://t.co/zoX0lkUGuo

[Xiaoyin Qu @quxiaoyin]

Prediction: from now on, great product leaders will get $100M comp package, not researchers.

1. 2022-2026 is all about killer models. Right now models still matter, but only for frontier tasks like science, math, trading etc. etc. You either kill cancer or you don't matter.
2. Instead, great products like muse, grokbot make AI truly accessible for people. The proliferation of AI matters way more now.
3. Open weight models make it cheaper and better, and the margin for most models will go down significantly, leading to less profits.
4. Training techniques are copyable, but unique data is not. All things equal(compute, training techniques), unique data is the new oil. Even if your training techniques are second-tier, with unique first-tier data you win.
5. Great products attract sticky users who provide unique data that lead to even better models.

Therefore, product is the one that matters. We are finally entering the gold era of product visionaries.

[MV @michaelvartikar]

funny to have this take after SWE already went thru this and it was ~fine

[corsaren @corsaren]

I give it like ~2 years before this starts doing really weird things to elite white collar entry-level hiring.

If you are a university underclassman, uhhhh maybe start getting used to that title.

[MV @michaelvartikar]

a product is what it’s like to use it

[MV @michaelvartikar]

so…more productive and empowered and far from being outright replaced? cool sounds great

[will depue @willdepue]

no investor, no startup guy, no counselor, no content creator, no musician will avoid going the way of the software engineer and mathematician