Yuchen Jin
@Yuchenj_UW
Cooking fun AI systems & products @databricks. Prev: co-founder & CTO @ Hyperbolic, OctoAI (acquired by @nvidia) Apache TVM, PhD @ University of Washington.
887 Following 308.9K Followers
AI bills are exploding everywhere.
Since joining Databricks, I've seen brilliant engineers across the industry (Distinguished Engineers, Berkeley/MIT/Stanford PhDs) spend months figuring out where to buy AI tokens.
Let me save you some time.
Choose Databricks AI Gateway:
- Fast open-source and proprietary models, available on Day 0.
- Smart Routing cut our internal coding costs by 35%.
- Easily track spending and set budgets across teams.
Spend less time managing AI tokens. Spend more time building.
Show more
“There is no moat.”
But moats are everywhere:
Ambition.
Taste.
Agency.
Curiosity.
Judgment.
Courage.
Obsession.
Persistence.
Integrity.
Patience.
This era belongs to those who refuse to be mediocre.
Show more
heard in Silicon Valley:
Security researchers will soon make more than AI researchers.
finally, OpenAI is doing something right.
The OpenAI vs. Anthropic subscription war is less about who has the best model, but more about who offers the most tokens per $.
Day 25/
We are doubling all usage limits for all Pro 100/200/500 tiers and Business Premium, effective today.
We heard your kind feedback and won’t be reducing usage limits for Pro 200, so your Work and Codex usage will stay the same. We’re also going to continue offering extended limits for dots for now, and usage of your dot will still continue to be separate from your default usage.
And we'll of course also keep shipping better models and other efficiency improvements that increase your effective usage over time.
Show more
Crazy convo with a friend today:
“Remember that super fast inference engine we built for image models?”
“Yeah, a year ago.”
“Dude, we wrote the WHOLE thing by hand.”
Holy shit. Just ONE year ago.
Show more
Ali and Sridhar (Snowflake CEO) are having their Kendrick vs. Drake moment on X.
To our friends at Snowflake: we'd love to onboard you to Unity Gateway and our AI products.
No hard feelings. But your latest DeepSeek model is 3 months old. We ship OSS models on Day 0.
Show more
According to Artificial Analysis, Gemini 4 Carbon is both smarter and cheaper than Opus 5.5 per task.
My only caveat: they benchmarked Gemini 4 Carbon using Antigravity CLI.
I haven't touched Antigravity in months, and I'm not eager to install it...
Will try it out in Codex and Omnigent to see how good it really is.
Show more
Are we entering another “Google is back” cycle?
Google’s new model Gemini 4 Carbon beats Opus 5.5 and GPT-6 Astra on most benchmarks and ranks #1# on the Artificial Analysis Intelligence Index v5.
Let’s see if it can shake up Anthropic’s IPO.
Show more
I'm so glad AI killed LeetCode-style coding interviews.
I've talked to young AI talent, and most would simply skip companies that still use LeetCode interviews.
I agree with Felix. Taste, empathy, principled thinking, business sense, and agency are what really matter.
Show more
The coding interview is dead, at least for my team at Anthropic. I've seen enough to know that it no longer provides signal. I removed it from loops this month.
I have a spicy, optimistic take - I think there's an opportunity for junior engineers:
Show more
The real 1-person billion-dollar company.
A solo smart, high-agency CEO + AI can have an outsized impact on the world.
Every day I ask myself: “Is my team AI-native enough?”
My rule: If AI can do it, humans shouldn't touch it.
AI already writes 95% of our code. But PRs still get stuck waiting for approvals.
I asked: “Do those reviewers actually read the code? If not, why not let an AI agent handle it?”
So we built one. Life immediately got better.
How fast an organization becomes AI-native will determine whether it thrives or dies.
Show more
2023: GPT-4 almost solved grade-school math
2024: IMO silver
2025: IMO gold
2026 May: Erdős unit-distance conjecture
2026 Sep: Navier–Stokes Millennium Prize Problem solved
2026 Oct: 722 AI-written math manuscripts
Today: OpenAI proves ETH is false
Tomorrow: P = NP?
Show more
“if you value intelligence above all other human qualities, you’re gonna have a bad time”
Aging like fine wine.
I previously argued that terminals and IDEs are the wrong interfaces for coding agents.
Phones used to be even worse for coding. Typing is slow, screens are tiny, and navigating a codebase is impossible.
But we’re no longer writing code by hand. So the iPhone Duo makes sense.
What I actually want isn’t a mobile IDE, but a control plane for a fleet of agents: delegating tasks, tracking progress, reviewing their work, occasionally intervening, and even watching them operate my remote computers.
I believe foldable phones can unlock an entirely new category of agent-native apps.
Show more
iPhone Duo is so damn sick!
“That means the 200+ neoclouds are going to see guppy-level death and consolidation.”
If neocloud stocks tank on Monday, I may have accidentally accelerated the guppy-level death...
Show more
People ask: “Are you dunking on neolabs?”
Most neolabs will fail. That’s normal. But I believe more AI breakthroughs are coming.
I’m skeptical of those following the same playbook as OpenAI and Anthropic. Raising $1B and spending 50% on GPUs doesn’t make you a $10B company.
I’m rooting for the contrarians. Challenge scaling laws. Challenge transformers. Challenge how models learn. Challenge the next-token prediction paradigm.
My bigger concern is that brilliant researchers don’t necessarily make great founders. In my experience, researchers often struggle to turn their research into great products / sustainable businesses.
Show more
Heard from a neolab friend:
“Even if we fail to train the best model, we locked in GPU capacity early. H100/B200/B300 rental prices have since gone up 2–3x. We can just resell the compute and print money.”
OK, so that’s the business model.
Show more
Elon just ruined the mood for all my neolab and neocloud friends out there.
Heard from a neolab friend:
“Even if we fail to train the best model, we locked in GPU capacity early. H100/B200/B300 rental prices have since gone up 2–3x. We can just resell the compute and print money.”
OK, so that’s the business model.
Show more
