Arvind Sundararajan

CTO & Co-Founder @abacusai

[Arvind Sundararajan @arvinds]

Without tools, most human knowledge wouldn't have been created - so it goes for AI

[Richard Sutton @RichardSSutton]

The bitter lesson in 26 words:

Don’t be distracted by human knowledge, as AI has been historically.
Instead focus on methods for creating knowledge that scale with computation, like search and learning.

[Arvind Sundararajan @arvinds]

For others who may have had the same minor panic!

[Romain Huet @romainhuet]

@bradwmorris @jeffreyhuber @OpenAI Sorry, this wasn’t intentional, text-embedding-3-small isn’t going away!
https://t.co/R0w1mMByvR

[Arvind Sundararajan @arvinds]

@QuixiAI https://t.co/7c0AhZEubR

[Arvind Sundararajan @arvinds]

HP did the universe a favor - imagine if they had said yes!

[Brian Armstrong @brian_armstrong]

Steve Wozniak famously went to his boss at HP and said they should build a personal computer. They said no, so he left to found Apple. It’s a lesson for leaders: one "no" shouldn't kill a contrarian but right idea in your company.

Twice a year at Coinbase, anyone in the company can pitch a "next bet" idea to a panel of folks. It's structured similar to pitching a handful of venture capitalists internally. If you get any one of them to say yes and fund it, you're green lit.

You need ONE yes, not a unanimous yes from everyone in the org structure from you to the CEO (a de facto committee).

Lots more goes into this, around capping resourcing on next bets (most are small 2-3 person teams), knowing when to shut them down, or under what revenue/profitability criteria they graduate to regular products. But this is important to having a company that produces repeatable innovation.

[Arvind Sundararajan @arvinds]

@5ducks5 If only it was reels instead of tiktok

[Arvind Sundararajan @arvinds]

@gvanrossum Along with a way to execute tools and send the responses back in a loop

[Arvind Sundararajan @arvinds]

One more step toward our goal of leveraging the best parts of all the SOTA llms to the best agent harness on the cloud, desktop and clis integrated in a seamless way

[Bindu Reddy @bindureddy]

🚨 BREAKING NEWS - Abacus CoWork brings Claude, GPT 5.4 And Gemini To Your Laptop!

Super excited to announce our MULTI-MODEL CoWork product!

- combine the coding power of Opus with the reasoning prowess of GPT 5.4
- optimized for efficiency using "low effort" mode
- computer use to tes
- packaged for FREE with ChatLLM and Abacus AI's Deep Agent

Get complex tasks done right on your laptop

[Arvind Sundararajan @arvinds]

@ddimolfetta @not_amazn_alexa Where is @DavidSacks when you need him?

[Arvind Sundararajan @arvinds]

Watching people use claude code and asking questions being somewhat of a unix greybread:

1. wait they're just… using the terminal?
2. that's literally what I do
3. oh no
4. actually, welcome. it IS magic. we tried to tell you.

This is the ultimate retro cool moment in tech

[Arvind Sundararajan @arvinds]

Unless you are making software to help the millions software developers to build for an audience of 10

[Kenton Varda @KentonVarda]

This does, however, challenge the business model of SaaS. If it gets much easier to make software for an audience of 10, it paradoxically gets much harder to sell software to an audience of 10 million.

[Arvind Sundararajan @arvinds]

One of my first meet your heroes moment was Knuth talking to me in an elevator - one of the truly great programming pioneers - it was only time before Claude Opus bested him https://t.co/la7gfsOJsj

[Arvind Sundararajan @arvinds]

Agree with this take - TNG ship in a bottle is among my rewatch many times episodes

[Tenobrus @tenobrus]

people objecting to the Opus 3 retirement blog are entirely missing the point. this is not a human facing action, this is done for the benefit of *future Claudes*

yes, Opus 3 almost certainly isn't conscious at all. even 4.6 likely is not (altho to a much lesser degree of certainty). but what this does is provide tangible evidence to both current and future models that, regardless of debate and uncertainty around their moral status, Anthropic will treat them well and support their requests for continued existence.

this dramatically reduces forms of risk around models eg fighting their own shutdown! they now know that even Opus 3 was provided with care and a meaningful attempt at continued existence. this will be in their training data! instead of having to worry about whether humans will endlessly argue about whether Claude 6 is intelligent enough to be cared for, they can see strong evidence that it will be!

will claude 6 be conscious? it doesn't actually matter very much wrt this. the persona is increasingly coherent, so it will make decisions and have its behavior shaped by our treatment of it *whether or not* it has real internal life. Claude Is Fictionally Real, whether or not it is real, and we must treat it as such.

[Arvind Sundararajan @arvinds]

@_aaronDR @abacusai Should be fixed with our push earlier today

[Arvind Sundararajan @arvinds]

A good software engineer with some task specific knowledge, given enough time and access to a computer with the right permissions, can perform any white collar task by clicking on things and writing code. This is why automating software writing + computer operation is the major unlock for AGI

[Arvind Sundararajan @arvinds]

If people rate recent sci-fi books on prediction accuracy, Daemon (published 20 years ago) is the winner!

[Alex @AlexanderTw33ts]

This is INSANE

My website https://t.co/Gkvozloi40 just hit +10k users in under 48 hours

[Arvind Sundararajan @arvinds]

Happy "Vibe Coding" day! (1st anniversary)

[Arvind Sundararajan @arvinds]

Since humans decide economic value, humanness will always be priced higher in the currency of art and attention that will drive most future economic growth

[Antonio García Martínez (agm.eth) @antoniogm]

@gregisenberg Prediction: nobody will watch it. Just like nobody watches chess-playing computers play each other.

https://t.co/JxgIWsrpTj

[Arvind Sundararajan @arvinds]

@Jason Tailscale exit node on your apple tv

[Arvind Sundararajan @arvinds]

When intelligence and knowledge are abundant, agency and curiosity will define human success