q: how to witness a lifetime worth of bullsh*t?
a: set up an indian legal entity.
the infrastructure is not built to support growth and no one cares enough to solve it cause at the end they get votes for religious polarization.
The smartest Indians continue to come to build in the US for the simple reason that the economic incentives are far far better.
Perplexity ($20B), Skild ($14B+), Baseten ($13B), Together ($8B), Glean ($8B), Modal ($5B), Rubrik ($25B), Harness ($5B), Blacksmith, Parallel,
Nvidia - making chips
Google - making chips
Meta - making chips
Microsoft - making chips
Amazon - making chips
OpenAI - making chips
Anthropic - making chips
Alibaba - making chips
Everyone’s pushing to vertically integrate chip design to reduce reliance on Nvidia. Customers
OpenAI had to create the market and fail at things so Anthropic didn’t. They struggled to get traction and capital initially, making the enterprise GTM a no brainer cause that’s what you do with less capital.
It’s funny how all this is just 3 years 😂
@jackson_felty@aloobhujiyan you can isolate them by operating dedicated cards that auto-scale. we do this all the time for our customers while maintaining single tenancy.
@Pushpakteja exactly what we spend all our time on solving. also everyone markets the p50. we optimize for the p90/95/99.
we will launch and go public with more of this soon.
fair point: all roads lead to inference if you're engineers building in infra.
however, with inference being such a large market with so many options, specializing towards a workload and offering a stack that improves with more telemetry, allows to build long term stickiness with an amazing nps within customers.
the entire stack gets tuned to your workload patterns (input, output, cache hits, etc) and over a period of time, re-doing the cycle with another provider is like switching your database.
specialize:
codings agents and voice agents need completely different inference infrastructure, not just kernels. baseten and fireworks have spent so much capital building infra for cursor, that they have made everyone believe that the only way to tune performance is custom kernels.
Inference is the second way to raise a massive round right now. After the labs, you can still justify massive rounds by pointing at data centers and compute, which means more inference companies keep getting funded even as the underlying product stays extremely easy to switch
@itskuldeeep depends on businesses. I’m sure sales teams will still scale by virtue of the sheer human relationships to manage on customer side.
for us, we have individual owners for each piece of the stack, shrinking it isn’t ideal.
AI has gotten so great!
our engineering team hasn’t grown at all in the last 12 months. just a team of cracked engineers with multiple claude subs and our in-house GLM fine tune.
in this period, we have gone from zero to millions in revenue.
we are serving 100x the requests as compared to our direct competitors.
key customers have also grown inference volumes by 3-5x within a quarter.
Executives at OpenAI, Anthropic and other AI companies are reportedly preparing for a public and political backlash after a catastrophic AI event, possibly as early as 2027 (via Axios).
Axios reports that many industry insiders expect a major incident within the next 6–12
80 Followers 665 FollowingMachine Learning at Google Search (AI Mode / AI Overviews), BTech IIT Madras. Quant Researcher at JP Morgan in my previous life
Views are my own
331 Followers 4K FollowingBoilermaker who loves walks, music and chai. Engineer who will invent a new cancer drug in his lifetime(wannabe oncologist !)
4.0M Followers 839 FollowingCo-founder & CEO @Coinbase. Creating more economic freedom in the world. Co-founder @researchhub @newlimit.
Not investment advice.
36K Followers 5K Followingtweets about AI and other fun stuff. currently @foundationcap; wrote the context graph paper.
previously McKinsey, @georgiatech, @stackfolio (acquired),
530K Followers 6K FollowingCo-founder & CEO of Discovery Loop. Former Chief Scientist, Google. Helped build many Google products, TPUs, Gemini, TensorFlow, MapReduce, Bigtable, ...
45K Followers 2K FollowingFounder, Mettle Capital @mettlecap. Early-stage investing in India / India-US.
Write about AI, Tech, and India. [email protected]