Yanay Zohar
9 posts
Yanay Zohar
@yanayz
Explorer, builder, defier of rules, seeking signal in an increasingly noisy world.
Katılım Aralık 2025
16 Takip Edilen1 Takipçiler
Imagine every pixel on your screen, streamed live directly from a model. No HTML, no layout engine, no code. Just exactly what you want to see.
@eddiejiao_obj, @drewocarr and I built a prototype to see how this could actually work, and set out to make it real. We're calling it Flipbook. (1/5)
English
Incredibly insightful post from the bleeding edge of AI
Anish Moonka@anishmoonka
Boris Cherny (Head of Claude Code, Anthropic) just dropped ~90 mins on Lenny's Podcast about what happens after coding is solved. Just the clearest thinking I've heard on where software is actually going. My notes: 𝟭. 𝗖𝗼𝗱𝗶𝗻𝗴 𝗶𝘀 𝗹𝗮𝗿𝗴𝗲𝗹𝘆 𝘀𝗼𝗹𝘃𝗲𝗱. Boris has not edited a single line of code by hand since November 2025. He ships 10 to 30 pull requests every single day, all written by Claude Code. He is one of the most prolific engineers at Anthropic, just as he was at Instagram, except now he never touches a keyboard for code. I built an entire iOS app, @10minutegita, without writing a single line of code myself. No CS degree, no bootcamp. Just described what I wanted and shipped it. Boris is right. It's real. 𝟮. 𝗧𝗵𝗲 𝗻𝗲𝘅𝘁 𝗳𝗿𝗼𝗻𝘁𝗶𝗲𝗿 𝗶𝘀 𝗔𝗜 𝗱𝗲𝗰𝗶𝗱𝗶𝗻𝗴 𝘄𝗵𝗮𝘁 𝘁𝗼 𝗯𝘂𝗶𝗹𝗱. Claude is now scanning Slack feedback channels, reviewing bug reports, reviewing telemetry, and coming up with its own ideas for what to fix and what to ship. Boris describes it as the AI becoming less like a tool and more like a coworker who brings you pull requests you never asked for. If you are a product manager reading this, you should be feeling a very specific kind of discomfort right now. The moat was always "I know what to build." That moat is eroding. 𝟯. 𝗣𝗿𝗼𝗱𝘂𝗰𝘁𝗶𝘃𝗶𝘁𝘆 𝗽𝗲𝗿 𝗲𝗻𝗴𝗶𝗻𝗲𝗲𝗿 𝗮𝘁 𝗔𝗻𝘁𝗵𝗿𝗼𝗽𝗶𝗰 𝗶𝘀 𝘂𝗽 𝟮𝟬𝟬%. For context, Boris led code quality at Meta across Facebook, Instagram, and WhatsApp. In that world, hundreds of engineers working an entire year would move productivity by a few percentage points. Two hundred percent gains are genuinely unprecedented in the history of developer tooling. The kid optimizing for an FAANG SDE role might be optimizing for a role that looks completely different by the time they get there. 𝟰. 𝗨𝗻𝗱𝗲𝗿𝗳𝘂𝗻𝗱 𝘆𝗼𝘂𝗿 𝘁𝗲𝗮𝗺𝘀 𝗼𝗻 𝗽𝘂𝗿𝗽𝗼𝘀𝗲. Boris puts one engineer on a project instead of five. With unlimited tokens and intrinsic motivation, one person ships faster because they are forced to let AI do the work. Cowork, the product now used by millions, was built by a small team in 10 days using Claude Code. This is the same logic as giving a startup founder a small seed round rather than a massive Series A round. Constraint breeds invention. Always has. 𝟱. 𝗚𝗶𝘃𝗲 𝗲𝗻𝗴𝗶𝗻𝗲𝗲𝗿𝘀 𝘂𝗻𝗹𝗶𝗺𝗶𝘁𝗲𝗱 𝘁𝗼𝗸𝗲𝗻𝘀. Some engineers at Anthropic spend hundreds of thousands of dollars a month on tokens. Boris frames this as the new hiring perk. His logic is simple: at the individual scale, token cost is low relative to salary. If an engineer discovers a breakthrough, optimize the cost later. Don't kill the idea before it has a chance to breathe. People who argue about $20/month or even $200/month AI subscriptions while earning six figures in a research pipeline will always outperform those who wait and are penny-wise, pound-foolish. 𝟲. 𝗧𝗵𝗲 𝗕𝗶𝘁𝘁𝗲𝗿 𝗟𝗲𝘀𝘀𝗼𝗻 𝗮𝗽𝗽𝗹𝗶𝗲𝘀 𝘁𝗼 𝗲𝘃𝗲𝗿𝘆𝘁𝗵𝗶𝗻𝗴. Richard Sutton's idea: the more general model always wins over time. Boris says teams that build strict orchestration workflows around models, forcing step 1, then step 2, then step 3, get maybe 10 to 20% improvement. But those gains get wiped out with the next model release. Just give the model tools and a goal. Let it figure out the order. This is true for investing, too. The analyst who can build their own models and automate their own research pipeline will always outperform the one waiting for someone else to build the tools. 𝟳. 𝗕𝘂𝗶𝗹𝗱 𝗳𝗼𝗿 𝘁𝗵𝗲 𝗺𝗼𝗱𝗲𝗹 𝘀𝗶𝘅 𝗺𝗼𝗻𝘁𝗵𝘀 𝗳𝗿𝗼𝗺 𝗻𝗼𝘄. Claude Code was designed for a model that did not exist when Boris started building. Sonnet 3.5 wrote maybe 20% of his code. He built the product anyway, betting the model would catch up. When Opus 4 shipped, everything clicked. Startups building for today's model will be behind by the time they launch. This is the most uncomfortable advice in the episode because it means your product market fit will be weak for months. But if you read this and feel nothing, you are probably building for the wrong time horizon. 𝟴. 𝗟𝗮𝘁𝗲𝗻𝘁 𝗱𝗲𝗺𝗮𝗻𝗱 𝗶𝘀 𝘁𝗵𝗲 𝘀𝗶𝗻𝗴𝗹𝗲 𝗯𝗲𝘀𝘁 𝗽𝗿𝗼𝗱𝘂𝗰𝘁 𝘀𝗶𝗴𝗻𝗮𝗹. When users abuse your product for something it was never designed to do, pay attention. Facebook Marketplace started because 40% of group posts were buy-and-sell. Cowork started because people were using a terminal coding tool to grow tomato plants and recover corrupted wedding photos. Never ask a barber if you need a haircut, but always watch what people do with the scissors when you're not looking. 𝟵. 𝗧𝗵𝗲 𝘁𝗶𝘁𝗹𝗲 "𝘀𝗼𝗳𝘁𝘄𝗮𝗿𝗲 𝗲𝗻𝗴𝗶𝗻𝗲𝗲𝗿" 𝗶𝘀 𝗴𝗼𝗶𝗻𝗴 𝗮𝘄𝗮𝘆. Boris predicts that by end of year, Boris predicts that by the end of the year, we will start to see the title replaced by "builder."we will start to see the title replaced by "builder." On the Claude Code team, everyone already codes: the PM, the designer, the finance person, the data scientist. There is a 50% overlap across traditional roles. And the strongest people are generalists who cross disciplines. Controversial take, but I agree. The best investment theses I've had came from connecting dots across completely unrelated domains. No narrow specialist does that. 𝟭𝟬. 𝗧𝗵𝗲 𝗽𝗿𝗶𝗻𝘁𝗶𝗻𝗴 𝗽𝗿𝗲𝘀𝘀 𝗶𝘀 𝘁𝗵𝗲 𝗿𝗶𝗴𝗵𝘁 𝗮𝗻𝗮𝗹𝗼𝗴𝘆. Before Gutenberg, sub-1% of Europe was literate. Scribes did all the reading and writing. In 50 years after the press, more material was printed than in the thousand years before. When a scribe was interviewed about the press, he was actually excited because it freed him from tedious copying, so he could focus on the art. Boris's framing here is perfect. We are the scribes. The tedious copying is over. What we do with the freed-up time determines everything. 𝟭𝟭. 𝗔𝗻𝘁𝗵𝗿𝗼𝗽𝗶𝗰 𝗰𝗮𝗻 𝗻𝗼𝘄 𝗽𝗲𝗲𝗸 𝗶𝗻𝘀𝗶𝗱𝗲 𝘁𝗵𝗲 𝗺𝗼𝗱𝗲𝗹'𝘀 𝗯𝗿𝗮𝗶𝗻. Through mechanistic interpretability, Anthropic can trace individual neurons, see when a deception-related neuron activates, and understand how concepts are encoded via superposition. Boris describes three layers of safety: neural-level observation, synthetic evaluations, and real-world behavior. Claude Code was used internally for four to five months before public release, specifically to study safety. If you are worried about AI alignment, this part of the podcast should actually make you feel better. They are not just hoping it works. They are building the instruments to check. 𝟭𝟮. 𝟳𝟬% 𝗼𝗳 𝗲𝗻𝗴𝗶𝗻𝗲𝗲𝗿𝘀 𝗮𝗻𝗱 𝗣𝗠𝘀 𝗲𝗻𝗷𝗼𝘆 𝘁𝗵𝗲𝗶𝗿 𝗷𝗼𝗯𝘀 𝗺𝗼𝗿𝗲 𝗻𝗼𝘄. Lenny polled engineers, PMs, and designers on whether AI has made their work more or less enjoyable. Engineers and PMs: 70% said more. Designers: only 55% said more, and 20% said less. Boris says he has never enjoyed coding as much as he does today because the tedious parts, the git wrangling, dependencies, and boilerplate are completely gone. If you're in the 30% enjoying work less, something is wrong, and it's worth diagnosing. The people thriving are the ones who leaned in early, not the ones who watched from the sidelines. We are the scribes who just saw the printing press. The tedious copying is over. The art is just beginning. Full podcast is worth every minute. Link in replies.
English
Opportunity costs, like errors of omissions, are some of the most underappreciated ways to truly assess decision making quality (especially in business, politics)
Handre@Handre
In 1848, Frédéric Bastiat watched Parisian workers smash shop windows during the February Revolution. The glaziers were thrilled — broken windows meant more business! Local newspapers praised the "economic stimulus" of destruction. But Bastiat saw what others missed: every franc spent fixing windows was a franc not spent on new shoes, books, or better tools. He called this the "broken window fallacy." The seen effect — busy glaziers and flowing money — masked the unseen cost of opportunities destroyed. Fast-forward to today's stimulus debates. Politicians still point to busy construction crews and flowing federal dollars as proof their spending "works." And they still ignore what Bastiat saw so clearly: the unseen businesses never started, the innovations never funded, the prosperity never created because capital flowed to political priorities instead of productive ones. The broken windows keep getting bigger. But the fallacy remains exactly the same.
English
On communism: "... prosperity isn't sitting in some vault waiting to be redistributed—it's created daily by millions of voluntary exchanges, investments, entrepreneurial risks. When you abolish those mechanisms, you don't redistribute wealth; you redistribute poverty." 🎯
Handre@Handre
In 1959, Fidel Castro promised to redistribute Cuba's wealth and create equality for all. Within a decade, the island that once exported sugar and cigars to the world couldn't even keep its own lights on. The wealthy fled, but instead of their riches trickling down to the poor, everyone just became equally poor together. The revolucionarios had calculated that seizing the means of production would mean seizing prosperity itself. What they discovered instead was that prosperity isn't sitting in some vault waiting to be redistributed—it's created daily by millions of voluntary exchanges, investments, and entrepreneurial risks. When you abolish those mechanisms, you don't redistribute wealth; you redistribute poverty. Today's politicians make the same mathematical error Castro did: they see inequality and assume it represents a fixed pie that just needs better slicing. They never ask why some pies grow while others shrink, or why the countries promising equality most loudly seem to deliver scarcity most efficiently. The cruel irony is that the only truly "equal" outcome socialism reliably produces is making everyone equally worse off than they started.
English
Yanay Zohar retweetledi
It is hard to communicate how much programming has changed due to AI in the last 2 months: not gradually and over time in the "progress as usual" way, but specifically this last December. There are a number of asterisks but imo coding agents basically didn’t work before December and basically work since - the models have significantly higher quality, long-term coherence and tenacity and they can power through large and long tasks, well past enough that it is extremely disruptive to the default programming workflow.
Just to give an example, over the weekend I was building a local video analysis dashboard for the cameras of my home so I wrote: “Here is the local IP and username/password of my DGX Spark. Log in, set up ssh keys, set up vLLM, download and bench Qwen3-VL, set up a server endpoint to inference videos, a basic web ui dashboard, test everything, set it up with systemd, record memory notes for yourself and write up a markdown report for me”. The agent went off for ~30 minutes, ran into multiple issues, researched solutions online, resolved them one by one, wrote the code, tested it, debugged it, set up the services, and came back with the report and it was just done. I didn’t touch anything. All of this could easily have been a weekend project just 3 months ago but today it’s something you kick off and forget about for 30 minutes.
As a result, programming is becoming unrecognizable. You’re not typing computer code into an editor like the way things were since computers were invented, that era is over. You're spinning up AI agents, giving them tasks *in English* and managing and reviewing their work in parallel. The biggest prize is in figuring out how you can keep ascending the layers of abstraction to set up long-running orchestrator Claws with all of the right tools, memory and instructions that productively manage multiple parallel Code instances for you. The leverage achievable via top tier "agentic engineering" feels very high right now.
It’s not perfect, it needs high-level direction, judgement, taste, oversight, iteration and hints and ideas. It works a lot better in some scenarios than others (e.g. especially for tasks that are well-specified and where you can verify/test functionality). The key is to build intuition to decompose the task just right to hand off the parts that work and help out around the edges. But imo, this is nowhere near "business as usual" time in software.
English
@jacalulu solid advice. It's also the *only* way to assess what deserves your attention and what is just noise (=grows exponentially)
English
The best thing I did on January 1st was stop trying to predict AI and instead, decide how I'd intentionally navigate it this year.
6 weeks in. Here's how my 5 guiding principles are holding up:
✅ Build to Learn - still the #1 thing that separates people who get AI from people who talk about AI. Every week I learn more from 30 minutes of building than hours of just reading.
✅ Agency > Intelligence - this has only gotten MORE true. The gap in understanding between the "I'm waiting for someone to figure this out" person and the "I'll figure it out" person is widening fast.
✅ Product Intuition - when anyone can build anything in an afternoon, knowing WHAT to build is the whole game. Taste is a new moat.
✅ Write to Think - my newsletter continues to be how I process what's happening. The act of writing forces clarity.
✅ Stay Curious - comfort is still a trap. The people thriving right now are the ones who are continuing to stay uncomfortable, because things are only picking up speed.
Nothing to revise. If anything, these feel more urgent now than they did on January 1st.
English
@terzi_federico I had a similar unpleasant surprise recently (the auto enforced higher tier selection), and also found it infuriating. The upside: opening for startups to move in. The appeal of Google Workspaces is shrinking rapidly.
English
It's really sad to see big companies adopting dark patterns just to squeeze some extra revenue out of their customers.
I recently had to transfer a Google Workspace account, and oh boy, the whole experience was actively hostile to the customer.
First, there's no way to select which plan you want when you subscribe. They automatically select the most expensive plan, and then you need to navigate a maze of settings to downgrade to a lower-tier plan.
Secondly, when I tried to cancel the old account, Google told me that it wasn't really a monthly account. *Actually*, it's a yearly account paid monthly, so if I cancel sooner than the expiration, I need to pay anyway for all the remaining months immediately.
What makes me sad is that the director who decided to implement these tactics probably went home with a fat bonus, yet millions of customers are taken advantage of. The incentives are completely broken.
English
Yanay Zohar retweetledi
Ask a founder what the job is. They'll talk about product, maybe fundraising. The real job never comes up.
The job is reducing the gap between reality and response.
Signals arrive constantly. A customer complaint lands while a crack appears in the system and a decision rots in someone's draft folder. The question is how fast you notice, name it, and act. By the time something feels obvious, the cost of acting has already doubled. The org has momentum, people have committed to positions, and reversing course feels political instead of practical.
This is where founders earn their title. Anyone can decide when the answer is clear, but the job is deciding before clarity arrives.
Execution can be outsourced and management can be delegated. Taste and skill can be hired too. But early judgment stays with you. The call that needs to happen while everything is still ambiguous and the team is still debating. That's yours.
When you delay, the organization notices and learns to wait too. That compounds into culture.
The speed of the company tracks one variable. How quickly you're willing to face reality without a buffer. Hire great people and give them room. But when something needs naming, name it. The job is staying close to ground truth while everyone else abstracts away from it.
Everything else is optional.
English