Pinned Tweet
Announcement: I'll be joining @DecagonAI tomorrow in San Francisco as an intern on the AI Infra team working on AI DevX!
At Decagon, I'll try to answer the question: "how can we build systems that improve the experience, productivity, and quality of AI use across a company?"
If I'd love to chat. Excited to work and see all of you soon!
At Decagon, I'll try to answer the question: "how can we build systems that improve the experience, productivity, and quality of AI use across a company?"
If I'd love to chat. Excited to work and see all of you soon!
35
1
136
9,865
Tyler Xiao retweeted
Devin can do more than ship PRs. @DecagonAI used it to run an inference research lab.
One Devin orchestrates, subagents chase competing hypotheses in parallel, and every idea is tested against replayed production traffic on @modal GPUs. The result was up to 2.5× more throughput per GPU.
One Devin orchestrates, subagents chase competing hypotheses in parallel, and every idea is tested against replayed production traffic on @modal GPUs. The result was up to 2.5× more throughput per GPU.
4
10
85
7,086
Tyler Xiao retweeted
I think many of us working in tech tend to forget that the reason software engineering vs other engineering fields pays (a lot!) more, is a lot more flexible (WFH/remote), and has better perks:
Because this field changes much faster than other engineering fields. The premium is because fewer people are able/willing to pick up new technologies, methodologies so quickly.
But the field had ~10-20 years of stability, and so lots of folks forgot this is why a premium exists. And that AI is just another tech wave: software engineers who pick it up and master it will generally fare better; those not will struggle a lot more.
Because this field changes much faster than other engineering fields. The premium is because fewer people are able/willing to pick up new technologies, methodologies so quickly.
But the field had ~10-20 years of stability, and so lots of folks forgot this is why a premium exists. And that AI is just another tech wave: software engineers who pick it up and master it will generally fare better; those not will struggle a lot more.
275
532
6,990
299,753
Tyler Xiao retweeted
marketing team advised that I shitpost less and try to provide some real value to the world, so here's what I've been working on recently! x.com/i/article/2112…
7
6
40
1,727
If you’re a founder or engineer tired of user bug reports and issues flooding your software feed, try @Simulithic and give these guys a quick call!
Shipping code shouldn’t mean hours of manual testing.
@Simulithic (YC F26) sends AI agents through your app like real users, across browsers, devices, and account types. It tells you what broke without you writing a single test script.
The teams that ship fastest will be the fast as they build.
Get started today: simulithic.com
1
0
12
748
.@bryan_johnson pay attention to how this guy Satyam de-aged these past years 🫡
Doing a hackathon with @eduardfausgil & @ayushrgarg was the best decision I’ve ever made
I was in India, about to give up on tech back in 2022, met these 2, learned how to build cool things with them, and it changed my life
(Picture below is approximately 4.5 years apart lol)
1
0
5
633
The lesson we have learned over the past few years is that frontier general intelligence is a game changer, but not a drop in replacement for complex, end-to-end real world tasks.
I still remember last year when OpenAI announced that its general purpose model solved the IMO Gold. Fast forward today and this level of intelligence/problem-solving is available with a prompt costing less than $20.
Even with this level of AI capability, the real world is messy. Building AI-native companies that produce value and solve real issues is still valuable these days :)
I still remember last year when OpenAI announced that its general purpose model solved the IMO Gold. Fast forward today and this level of intelligence/problem-solving is available with a prompt costing less than $20.
Even with this level of AI capability, the real world is messy. Building AI-native companies that produce value and solve real issues is still valuable these days :)
value accrual at the model layer is splitting in two. for frontier discovery, it concentrates in a handful of labs. for commoditized workflows, it leaks out to whoever is cheapest or closest to the customer.
the reason is that the value of being one model generation ahead varies enormously by task. for problems with known answers, intelligence eventually becomes abundant. for problems without known or verifiable answers yet, the strongest model can be worth orders of magnitude more than the next best.
frontier discovery is novel research and long-horizon agentic work where nobody knows the answer yet. bio appears to be a large one over the next few years for ex. models are becoming exceptionally strong at prediction, and the value of a single hit dwarfs the cost of the tokens that found it. expect the labs to start running their own experiments, with frontier models choosing what to test. the feedback loop here isn’t just better models producing better predictions. it’s better predictions determining which experiments get run, generating new ground truth that makes the next round of predictions better.
openai and anthropic are best positioned here, with google close behind. buyers of discovery pay for the strongest model, so being a generation behind costs far more than it does anywhere else. the lead compounds given the labs use their best models to accelerate building the next ones, they have the compute to do it at scale, and they keep those models internal long before exposing an api.
commoditized workflows are the long tail of tasks a model one generation old already handles well. quality saturates and cost decides. deepseek and other chinese labs tend to own the cost-quality pareto frontier, partly out of necessity given their relatively compute-constrained conditions, meaning efficiency has become integral to the research agenda. the western frontier labs have their own edge in distillation, as they can leverage the strongest teachers at the logit level, while everyone else distills from a weaker teacher or from sampled text. but they only win this tier if they choose to compete on price.
even the pareto leader doesn’t capture much per task here, because the weights are generally open. anyone can self-host them or finetune them on their own data, and as a result the model layer’s share shrinks to an inference margin. but thin margins on an enormous volume of workflows can still add up to substantial value capture. the model layer becomes a scale business rather than a pricing power business.
cost isn’t the only place value moves. whoever owns the customer relationship also owns the workflow, the context, and increasingly the feedback that determines what good performance actually looks like.
a caveat then is taste. general taste lives in the weights, but most “commodity” workflows are full of judgment calls specific to one business. the ground truth for those comes from outcomes the labs never see. if you’re building on top of the labs, don’t compete on general intelligence. own the loop that turns your customers’ outcomes into a reward signal.
3
1
19
1,859
Tyler Xiao retweeted
The last remaining undefeated team in the Big Ten or SEC is UCLA.
106
814
17,517
496,033
Tyler Xiao retweeted
the builder’s lounge for @vestucla @cognition @warpdotdev and rainfall ventures for @Techweek_!!
thank you to everyone who helped make this happen! we had a packed house of students, founders, and people across the LA startups space in attendance and a waitlist or over 400!
i this with the team and can’t wait to do more events! if you’re looking to meet people or host with people in the LA tech space, dm me!
thank you to everyone who helped make this happen! we had a packed house of students, founders, and people across the LA startups space in attendance and a waitlist or over 400!
i this with the team and can’t wait to do more events! if you’re looking to meet people or host with people in the LA tech space, dm me!
7
4
29
853
