Samarth Jain

connecting dots others haven't drawn yet

[Samarth Jain @SamarthJainnn]

Been thinking this for a while: What if the best brain-sensing wearable is already in your ears? We wear AirPods all day, right next to the brain. Apple quietly patented EEG AirPods in 2023, and NextSense and IDUN already ship in-ear EEG. The future of neural tech might not be a new gadget. It might be the one you forgot you were wearing. Really curious when we actually get to see this in our own hands.

[Samarth Jain @SamarthJainnn]

i swear claude just deleted my chat out of nowhere rn

[Samarth Jain @SamarthJainnn]

couldn’t have put it better @prvkhvr ! GenZ derives its sense of community from the internet, and has grown up in this sort of an ‘interest economy’.

[Prakhar Gupta @prvkhvr]

This alone will help you understand all gen z discourse.

[Samarth Jain @SamarthJainnn]

Two models, same company, opposite answers – and the wrong one sounded more confident.

[Samarth Jain @SamarthJainnn]

Two models, same company, opposite answers – and the wrong one sounded more confident.

[Samarth Jain @SamarthJainnn]

Fable 5 dropped, so I did a fun experiment: asked two Claude models the same question — "what is Fable 5?"
Sonnet gave me a confident, detailed breakdown: persona prompts, structured inputs, use-case framing. Sounded completely plausible. Then I asked Fable 5 itself. Totally different answer.

[Samarth Jain @SamarthJainnn]

Fable 5's version: "Sonnet made that up. I launched after its training cutoff – it pattern-matched your question and confabulated."

[Samarth Jain @SamarthJainnn]

So pricing was never capped by open source. It's capped by physics (megawatts, transformers, fab slots). Open or closed, subsidized or metered, every model on earth bottlenecks at the grid. Access to power, not algorithms, is the binding constraint. The only moat that doesn't evaporate when the next model drops is the one made of atoms.

[Samarth Jain @SamarthJainnn]

The closed vs open AI debate is happening one floor too high.
Both run on the same substrate: compute. Every token (Claude's, GPT's, or a free open model on OpenRouter) is a GPU burning power. The model is just software, and software needs silicon. The real bottleneck sits underneath the layer everyone's arguing about.

[Tommy @Shaughnessy119]

The most basic way AI could blow up imo. I'm not saying it does but this is the most obvious way I can see it happening

- Per seat subscriptions are massively subsidized. The flat fee was priced way below what heavy usage actually costs

- For real business use you have to move to the API anyway. Data protections, work integrations and compliance officer approval

- On the API you pay metered rates, and businesses are burning credits way faster than the per seat pricing ever led them to expect

- This is everywhere right now. Internally for us, Codex users, Uber torching its entire 2026 AI budget in 4 months, the Microsoft comments. Just go try an API

I shared more on this here: https://t.co/iZrqrCAIRW

- And I don't think most businesses have the money to keep paying increasing API rates without a real change to how they operate (caps needed)

- Because they have a cheap alternative. They can reach open source models through any aggregator (OpenRouter, Venice, Baseten, Together) and still get strong privacy. Venice private data centers, or E2EE/TEE serving GLM 5.1.

More on open source inference provider raises here: https://t.co/7kf56P44yQ

- And the discount is enormous. DeepSeek V4 codes within a hair of Opus on SWE bench at roughly 1/30th the price, and the cheapest open models run closer to 1/100th

- Chinese labs open source frontier grade models. The model is the single biggest cost an inference provider has, and they get it for free

- This idea dies if China goes closed source. That is actually bullish web2 AI labs, because if everyone is closed you pay up for the best intelligence. China goes closed source if they are tired of giving away an asset and they want the revenue and data flow to train new models

- Is this showing up in web2 AI lab revenue yet? No. Revenue is off the charts. Anthropic went from 9B to 47B run rate in five months

- So go forward, what happens?

- I think revenue slowly starts leaking to the open source inference providers (see Venice usage, OpenRouter's $113M raise, Baseten is raising at $11B or triple its valuation in three months, on revenue that went from $200M to $600M annualized in a single quarter)

- It doesnt move overnight, but it caps the labs ability to raise prices, and margins are already deeply negative. OpenAI is reportedly running near negative 122%

- With margins that bad there is no cash flow, so the labs are fully dependent on outside capital to buy GPUs, train models, and keep subsidizing usage (I.e. see Google tapping $80b equity sale, granted 30b for employee RSU taxes. Clearly they think Equity is overvalued or you wouldn't sell it)

- The break comes when that capital stops. Pricing is capped so margins cant improve, and the moment investors lose conviction on payback, the whole flow reverses

- Why would they lose conviction on payback? Back to the start - the inability to improve margins or get businesses to pay more

- This is also limiting, if we start making new drugs with AI or create entirely new businesses, you better believe people will pay up to the max for AI usage

[Samarth Jain @SamarthJainnn]

But even Nvidia isn't the real ceiling. You can build more chips in a year or two. You can't build more power that fast. AI data centers are on track to eat up to 12% of all US electricity by 2030, the giant transformers that feed them now take 3–4 years to get, and the wait to even plug into the grid runs for years. The squeeze isn't the computers anymore – it's the electricity to run them.

[Samarth Jain @SamarthJainnn]

Fully agree. Robotics is the true next frontier. We’re moving from digital AI to physical AI, and it just clicks for everyone. No one needs to explain what a robot doing a task means, it’s instantly intuitive. The demand is massive and the unit economics are insane. This wave is going to create trillion-dollar companies while replacing huge amounts of manual labor. The employment impact is going to be real and it’s coming faster than most people think.

[Samarth Jain @SamarthJainnn]

India is importing its present and risking its future. 89% of energy, nearly all its gold, its entire AI stack – all from abroad. On sovereign supply:

[Samarth Jain @SamarthJainnn]

https://t.co/fl4K8uRirr

[Samarth Jain @SamarthJainnn]

https://t.co/fl4K8uRirr

[Samarth Jain @SamarthJainnn]

update: these fake american air bots are spamming me nonstop. smh

[Samarth Jain @SamarthJainnn]

Sitting at PHX. @AmericanAir delayed my 8:15pm flight to SFO until midnight. Reason? ATS. A system-level failure that's nobody's fault, technically. But here's the thing: the FAA just admitted it has ~1,500 fewer controllers than needed. That's not weather. That's a chronic, documented infrastructure crisis passengers keep absorbing silently, with zero compensation.

[Samarth Jain @SamarthJainnn]

I have no idea how they came to the conclusion to vote out Moxie of all people.

[delian @zebulgar]

We put 12 of our favorite people in a room

@sama @zoink @bryan_johnson @PalmerLuckey @cyantist @traestephens

And made them play the diabolical game of deception, Mafia

And filmed it all live with @micsolana as the moderator

Good fun :)

[Samarth Jain @SamarthJainnn]

In today’s day and age when ideas have become cheap and execution has become easy, distribution is the true moat

[Samarth Jain @SamarthJainnn]

OpenAI x YC: @sama offering $2M in OpenAI tokens to every YC Startup in the current batch in exchange for equity is a very clever move for OpenAI:
- Low-cost currency with high perceived value (tokens have high retail pricing but OpenAI's marginal inference costs are much lower)
- Vendor Lock in and Switching costs (startups that heavily use the credits will build workflows, agents, evals, and fine-tunes around OpenAI's APIs- switching costs would be too high)
- Equity upside on the best outcomes (OpenAI gets a small stake in hundreds of promising companies without writing big cash checks)
- Enterprise Flywheel (Many YC companies scale into enterprises or get acquired by them. Early adoption here seeds OpenAI deeply into the next generation of B2B tools, agents, and workflows)