科技推文精选 - 2026年5月25日

2026年5月25日科技每日简报

Today's top tech conversations are led by @ShishirShelke1, whose post about 'A Chinese startup called Petti...' garnered the highest engagement. Key themes trending across the top stories include claude, every, because, engineers, company. The community is actively discussing recent developments in AI, engineering practices, and startup strategies.

1. ShishirShelke1 (Group Score: 166.0 | Individual: 30.4)

Cluster: 7 tweets | Engagement: 46 (Avg: 433) | Type: Tech

A Chinese startup called PettiChat has built an AI pet collar that claims to translate barks and meows into human language in real time.

It costs $118 and has reportedly received over 10,000 pre-orders already.

How tf is this possible? 💀 https://t.co/24cJahN3Tp\n\nQT @Polymarket: NEW: Chinese AI pet translating startup claims it can interpret pets' speech with up to 95% accuracy.

See 6 related tweets

@shiri_shh: chinese startup built an AI collar that translates barks and meows into full sentences.

95% accurac...

@FirstSquawk: CHINESE AI STARTUP CLAIMS IT CAN TRANSLATE PETS’ SOUNDS WITH UP TO 95% ACCURACY....
@cgtwts: We might actually be entering the era of talking to pets.

A Chinese AI startup says its collar can ...

@Cointelegraph: ⚡️ NEW: A Chinese startup says its AI-powered pet collar can translate barks and meows into full sen...
@VaibhavSisinty: Wait , WHAT!!!!

This Chinese AI startup claims it can interpret pets' speech with up to 95% accurac...

2. sandygrains (Group Score: 154.9 | Individual: 36.4)

Cluster: 5 tweets | Engagement: 7642 (Avg: 3147) | Type: Tech

RT @Ric_RTP: Microsoft just banned its own engineers from using AI.

The tool was literally costing MORE than the humans it was supposed to replace.

They lied to you about AI adoption and now the whole narrative is blowing up:

Microsoft gave thousands of engineers access to Claude Code six months ago and encouraged them to use it.

Engineers loved it and adoption exploded. But then the invoices arrived.

Token-based pricing means every query, every code review, every debugging session costs money. At scale across 100,000 engineers, the numbers became so large that Microsoft issued an internal order to cancel nearly all Claude Code licenses by end of June and force everyone onto their own cheaper tool instead.

The company that invested $5 billion in Anthropic just told its own people to stop using Anthropic's product because it costs too much.

Uber's story is even worse...

Their CTO Praveen Neppalli Naga told The Information that the budget he planned for the full year was "blown away already" by April.

Uber had rolled out Claude Code in December 2025. By March, 84% of their 5,000 engineers were using it with 70% of all committed code coming from AI systems.

Heavy users were burning $500 to$ 2,000 per month each. Naga himself spent $1,200 in a single two-hour demo session.

The company had even built internal leaderboards ranking engineers by how much AI they used. They literally gamified the spending and then ran out of money.

Now look at what Nvidia's own VP of applied deep learning Bryan Catanzaro said to Axios last month. Direct quote:

"For my team, the cost of compute is far beyond the costs of the employees."

This is a VP at the company that SELLS the chips saying that using AI is more expensive than paying humans.

Think about what this means for the entire AI narrative.

Every CEO on every earnings call for the past two years has said the same thing:

AI will make us more efficient, reduce headcount, and cut costs.

The stock market rewarded every company that said it.

Fired workers, stock goes up. Announced AI adoption, stock goes up.

But the actual companies deploying AI at scale are discovering the math doesn't work. The MORE employees use AI, the HIGHER the bill.

Goldman Sachs forecasts a 24x increase in token consumption by 2030 as companies adopt AI agents. Gartner just published a report showing that even though individual token prices will drop 90% by 2030, total enterprise AI costs will go UP because agents consume exponentially more tokens per task than basic tools.

Meta built an internal dashboard called "Claudeonomics" to track which employees use the most AI. Amazon started pushing engineers to "tokenmaxx," their internal term for consuming as many AI tokens as possible.

Both companies are spending hundreds of billions on AI infrastructure this year alone.

And Microsoft, the company that bet its entire future on AI, just told 100,000 engineers to stop using the tool they liked best because the per-token bills got out of control.

The companies building AI are telling investors it saves money. The companies using AI are finding out it costs more than the humans it was supposed to replace. And even the company that makes the chips just admitted it through its own VP.

This is the gap nobody on Wall Street is pricing in.

$725 billion in AI infrastructure spending this year across Big Tech. And the first companies to actually deploy these tools at scale are already pulling back because the economics don't work.

What do you think?

See 4 related tweets

@alex_prompter: Microsoft didn't cancel Claude Code because it was bad.

They canceled it because it was TOO GOOD.

...

@HedgieMarkets: 🦔Tech companies that pushed employees to maximize AI usage are now realizing the math does not work....
@alex_prompter: Marc Andreessen admitted on Joe Rogan that AI is making people less efficient.

The guy who funds ha...

@ataiiam: The conversation around demand for software engineers consistently misses a huge factor: the entire ...

3. danshipper (Group Score: 142.3 | Individual: 39.3)

Cluster: 5 tweets | Engagement: 125 (Avg: 105) | Type: Tech

If you read my essay After Automation this is the interview where I make practical predictions about what it means for your work:\n\nQT @lennysan: Automation is a lie. CLIs are over. The SaaSpocalypse is dumb.

A year ago @danshipper came on the podcast to predict where AI was heading. He was remarkably right—including the call that everyone was sleeping on Claude Code.

Dan has a unique lens into where things are going because his team at @every is possibly the most AI-pilled group of people in tech. I always learn a ton talking to Dan.

So I brought him back for round two. We'll score these in exactly a year: 🔸 Every company will have one “super-agent” in Slack. 🔸 Codex and Claude Code will become the new operating system for knowledge work. 🔸 The AI job apocalypse is not happening. 🔸 PMs and designers will thrive. 🔸 We will read way more AI-generated writing and we will like it. 🔸 "I would buy SaaS stocks right now."

Listen now 👇 https://t.co/wzxQ5bz49h

See 4 related tweets

@lennysan: .@danshipper: "I would buy SaaS stocks right now. SaaS stocks will be up majorly in the next couple ...
@clairevo: Two faves, back at it again.\n\nQT @lennysan: Automation is a lie. CLIs are over. The SaaSpocalypse ...
@bran_don_gell: Automation is a lie. Every agent needs a human.

If you’re bearish on humans, listen to this and su...

@every: RT @lennysan: Automation is a lie. CLIs are over. The SaaSpocalypse is dumb.

A year ago @danshipper...

4. minchoi (Group Score: 137.6 | Individual: 32.0)

Cluster: 5 tweets | Engagement: 254 (Avg: 168) | Type: Tech

Mythos release feels close.

But the real story is not the model drop.

It's what happens after:

10,000+ high/critical bugs found.

Maintainers asking Anthropic to slow down.

Finding bugs is now AI-speed. Patching is still human-speed.

What breaks first? https://t.co/PwMuTltOuN\n\nQT @testingcatalog: ANTHROPIC 🔥: Mythos 1, "claude-mythos-1-preview", is being prepared for a release on Claude Code and Claude Security.

The model became visible for a short amount of time on Claude; besides that, new strings mentioning Mythos have been added.

Access to the Claude Mythos model in Claude Code and Claude Security.

It still doesn't mean the general public will have access to this exact model, according to Anthropic's earlier communication.

More below 👇

See 4 related tweets

@kimmonismus: "We look forward to making Mythos-class models available through general release"

I don't understan...

@aiedge_: Claude Mythos spotted for a brief second in Claude!

Rumor has it that V1 of Mythos is being prepped...

@Kyrannio: Interesting\n\nQT @testingcatalog: ANTHROPIC 🔥: Mythos 1, "claude-mythos-1-preview", is being prepar...
@minchoi: RT @minchoi: Mythos release feels close.

But the real story is not the model drop.

It's what happe...

5. rohanpaul_ai (Group Score: 134.5 | Individual: 52.3)

Cluster: 3 tweets | Engagement: 288 (Avg: 50) | Type: Tech

🇨🇳 🇺🇸 China's Huawei’s new 122TB SSD shows how export controls can move innovation sideways instead of simply stopping it.

Huawei just built a 122.88TB AI SSD by changing the package around the memory, not by matching Samsung’s most advanced 400+ layer 3D NAND.

And a 245TB version discussed as a future step.

High-capacity SSDs usually grow by stacking more NAND layers inside each chip, but Huawei’s access to those chips is blocked because its Entity List status restricts items tied to US technology.

So it is not trying to win only by making taller 3D NAND stacks, where Samsung has already shown 400-plus-layer V-NAND work.

Instead, Huawei is shifting the contest from the chip itself to the way chips are packed together.

Huawei’s workaround is Die-on-Board, which puts NAND dies directly onto the circuit board, cuts out some normal chip packaging, and raises board-level density by packing more lower-density memory into the same device.

Direct die placement creates heat and signal problems, but it shows how packaging can recover some of the capacity lost when a company cannot buy the best memory chips.\n\nQT @rohanpaul_ai: Great article here on DeepSeek.

Their real story is not cheaper chatbots, but architecture that turns hardware scarcity into strategy.

DeepSeek is not trying to sell coding seats, it is trying to make Chinese memory, accelerators, and systems useful for frontier AI.

Every recent DeepSeek move attacks a bottleneck that makes frontier models dependent on elite HBM-heavy GPU stacks: MoE activates only parts of a model, DSA reduces long-context attention cost, and V4-Pro’s official card says CSA/HCA cuts 1M-token single-token inference FLOPs to 27% and KV cache to 10% of V3.2.

Engram, a separate research line, pushes the same logic from another side: let static knowledge live in scalable lookup memory, then fetch it predictably from host memory instead of forcing every fact through dense computation.

That sounds like engineering detail until you see the business consequence.

If models need less HBM and less brute-force compute, then second-best chips, abundant LPDDR, NAND, and customized ASICs become less second-best.

Reuters has already reported a permanent 75% DeepSeek V4-Pro price cut, while noting Huawei Ascend supply constraints and expected supernode availability, which is exactly the kind of feedback loop that they wanted.

DeepSeek is not only optimizing models for benchmarks, it is optimizing AI for a different industrial base.

The prize is not the app layer.

The prize is making scarcity programmable.

See 2 related tweets

@rohanpaul_ai: Great article here on DeepSeek.

Their real story is not cheaper chatbots, but architecture that tur...

@rohanpaul_ai: Reuters: DeepSeek just made its V4-Pro price cut permanent, pushing the price down to 25% of its ori...

6. levie (Group Score: 110.4 | Individual: 42.2)

Cluster: 3 tweets | Engagement: 4784 (Avg: 1520) | Type: Tech

CEOs are uniquely prone to AI psychosis because they’re sufficiently distant from the last mile of work that still has to happen to generate most value with AI.

So when they play with AI, they see the happy path results, often not considering the next 10 or 20 things that have to happen to get sustainable results from agents.

“Look I made this awesome product prototype”. Yes but you didn’t have to review the code before it went into production and fix a bunch of issues.

“Look I generated a contract”. Yes but you didn’t verify all the terms before it goes out to the counterparty and didn’t have to wire up all the past contracts to work with.

The best thing you can do as a CEO is to use AI a ton to figure out the real implications of agents in the enterprise, and come out the other side with an appreciation for both the upside and the real work that goes into them.\n\nQT @michalmalewicz: CEOs are the most delusional about AI. Detached from reality.

See 2 related tweets

@ShriramKMurthi: RT @levie: CEOs are uniquely prone to AI psychosis because they’re sufficiently distant from the las...
@jerryjliu0: I asked my eng team if I could ship code to prod

They told me no 💀\n\nQT @levie: CEOs are uniquely ...

7. mark_k (Group Score: 109.3 | Individual: 30.2)

Cluster: 4 tweets | Engagement: 299 (Avg: 191) | Type: Tech

Grok Build by @xai is now available for all SuperGrok subscribers 🔥

This is xAI’s agentic coding CLI, running directly from your terminal. Great to see access expanding beyond Heavy, because the AI coding race just got a lot more interesting.

Install:

"curl -fsSL https://t.co/E3Ero1CUbp | bash"

See 3 related tweets

@MarioNawfal: 🇺🇸 @Grok Build dropped for all SuperGrok users.

One line in your terminal and you're building auton...

@VaibhavSisinty: xAI just opened Grok Build up to all SuperGrok subscribers.

If you have SuperGrok you can now insta...

@MarioNawfal: RT @MarioNawfal: 🇺🇸 @Grok Build dropped for all SuperGrok users.

One line in your terminal and you'...

8. Polymarket (Group Score: 108.4 | Individual: 33.3)

Cluster: 6 tweets | Engagement: 9363 (Avg: 1551) | Type: Tech

NEW: AI is reportedly pushing McKinsey & rival consulting firms to rethink pricing, as clients are “questioning the value” of human advice.

See 5 related tweets

@edzitron: McKinsey is a company built entirely on grifting people by ingratiating their ideas in a very expens...
@Hesamation: Consultants charged $200K for a report ChatGPT could write using Deep Research and now can’t consult...
@EHuanglu: AI can analyze more data in 10 seconds than most consulting firms do in 100 days yet people still li...
@ShanuMathew93: McKinsey bruvs looking at the hyper capable agent they trained that is reducing their billable rate ...
@FT: How AI is forcing McKinsey and its peers to rethink pricing https://t.co/RvYI9GQ6Bz | opinion...

9. MTSlive (Group Score: 101.5 | Individual: 48.7)

Cluster: 3 tweets | Engagement: 3005 (Avg: 134) | Type: Tech

SITUATION DETECTED: Google DeepMind’s AI agent autonomously solved 9 of 353 open Erdos problems in mathematics, at a cost of a few hundred dollars per problem.

See 2 related tweets

@mark_k: Google DeepMind just dropped one of the clearest signals yet for where math is going.

The @GoogleDe...

@kimmonismus: Nine more Erdős problems have been solved.

This time, however, by Google DeepMind.

This shouldn't ...

10. imsroch (Group Score: 88.3 | Individual: 35.1)

Cluster: 3 tweets | Engagement: 622 (Avg: 154) | Type: Tech

RT @leerob: You might believe you should spend less time thinking about code because of AI.

I strongly disagree! We’re watching this play out live where tons of AI generated code becomes a liability.

At the end of the day, an engineer needs to be responsible / on call for code that gets shipped to production. If you don’t understand the system you’re trying to debug, you’re probably going to have a bad time.

Yes, AI can help with all of this, if you set up the proper systems. You can have agents triage prod logs, look at errors, etc. You can speed up parts of the investigation, but an engineer needs to make the call. There might be serious customer or financial implications from that change.

I expect the trend continue for trimming dependencies, vendoring code so you can modify it directly, preferring simpler systems with fewer abstractions, and spending waaaay more time thinking about system design and code maintenance.

I’ve said this before, but it’s a great time to get familiar with CS fundamentals and some of the history behind what great software looks like. Many parts will be different in the coming years as AI progresses, but also a lot more than people realize will stay the same.

See 2 related tweets

@badlogicgames: many such voices recently.\n\nQT @leerob: You might believe you should spend less time thinking abou...
@_philschmid: I am reading and looking easily 20x more at code than before AI. It is fun. And there wasn’t a bette...

11. Av1dlive (Group Score: 87.9 | Individual: 39.1)

Cluster: 3 tweets | Engagement: 142 (Avg: 56) | Type: Tech

Jensen Huang (CEO of Nvidia):

"I will always hire the graduate who's expert in AI over the one who isn't."

He said the quiet part louder

"Every college student should graduate as an expert in AI."

Because if your job can be automated, you are going to be disrupted.

The people who learn how to use AI with 100x more speed and efficiency are the ones who will stay ahead.

And he's not talking about people who occasionally ask Codex a question.

[He means people who actually understand how these systems work.]

→How to build with them. →How to automate workflows. →How to turn ideas into products faster than everyone else.

I used Claude Code for 12 months.

Then I switched to Codex for 30 days.

I realised most people are barely scratching the surface of what it can actually do.

So I put together a full Builder’s Guide on how to actually master Codex in 2026.

Read it and you'll already know more than 99% of people using it every day.

Full guide in the article below.\n\nQT @Av1dlive: https://t.co/pkEB9msLcW

See 2 related tweets

@Av1dlive: RT @Av1dlive: 2 OpenAI engineers just gave a masterclass on how to build and ship apps using Codex

...

@Av1dlive: RT @Av1dlive: Jensen Huang (CEO of Nvidia):

"I will always hire the graduate who's expert in AI ove...

12. chddaniel (Group Score: 85.8 | Individual: 30.2)

Cluster: 3 tweets | Engagement: 22 (Avg: 20) | Type: Tech

so you're telling me Claude Code Opus 4.7 can now...

scan an entire website
build it as a mobile app
prepare for App Store submission
self-maintain the app

without any human in the loop?!?

it's so over... https://t.co/pGGlyh9gzf\n\nQT @chhddavid: Introducing Website to App. Turn any website into an native mobile app.

Just paste a URL.

Claude Opus 4.7 will code, design, launch and translate a mobile app inspired by the original website.

We’ve been using this internally a ton for iOS/Android apps. https://t.co/hx4NLUeZ3u

See 2 related tweets

@chhddavid: this is f*king scary........\n\nQT @chhddavid: Introducing Website to App. Turn any website into an ...
@chddaniel: this is like actually terrifying...\n\nQT @chhddavid: Introducing Website to App. Turn any website i...

13. reach_vb (Group Score: 85.5 | Individual: 55.3)

Cluster: 2 tweets | Engagement: 2139 (Avg: 256) | Type: Tech

UPDATE: Came up with an even better version of this prompt after the feedback

Ask Codex to look across your sessions, Memories, and Chronicle, identify patterns, reuse what already exists, and only create the smallest useful skill, subagent, or automation.

"Look back over my recent work from the last 30 days, or all available history if shorter, and identify repeated manual workflows worth packaging.

Use available evidence in this order:

Recent Codex sessions and task summaries.
Codex Memories and rollout summaries to find patterns repeated across sessions.
Chronicle, if enabled, to spot repeated work outside Codex. Use Chronicle for discovery only; confirm important details in the relevant source system when possible.
Existing skills, custom agents, and automations, so you reuse or extend what already exists instead of duplicating it.

Look broadly for work that is repeated, time-consuming, error-prone, context-heavy, or benefits from a consistent process. Include workflows across coding, research, writing, planning, communication, operations, analysis, and personal administration.

Only act on a candidate when it:

occurred at least twice, or is clearly likely to recur and costly to repeat;
has stable inputs, a repeatable procedure, and a clear output or stopping condition;
would materially improve speed, quality, consistency, or reliability;
is not already adequately covered.

Choose the smallest appropriate form:

Skill: a reusable workflow or playbook.
Custom subagent: a bounded specialist role or investigation task suitable for delegation.
Automation: a scheduled or recurring check, report, reminder, or monitor.
Skip: work that is too one-off, ambiguous, sensitive, or poorly evidenced to package.

First produce a compact shortlist with:

repeated workflow
supporting evidence and dates
frequency/confidence
recommended form: skill, subagent, automation, extend existing, or skip
why it is or is not worth creating

Then create only the high-confidence missing items. Keep them narrow, practical, source-aware, and easy to validate. Do not create speculative, overlapping, or overly broad assets.

Finish with:

what you created or extended
what you deliberately skipped
what needs more evidence before packaging"\n\nQT @reach_vb: Copy and paste this into your codex:

“Look through my recent Codex sessions and identify repeated workflows or repeated asks.

For anything I keep doing manually, suggest:

a skill if it is a reusable workflow
a custom subagent if it is a bounded role or investigation task

Focus on practical things like CI failures, PR reviews, changelogs, docs updates, release prep, debugging, and test triage.

Create the useful ones only. Keep them simple.”

See 1 related tweets

@Dimillian: VIP. Very Important Prompt.\n\nQT @reach_vb: UPDATE: Came up with an even better version of this pro...

14. garrytan (Group Score: 75.4 | Individual: 38.7)

Cluster: 2 tweets | Engagement: 793 (Avg: 319) | Type: Tech

The companies I love working with in office hours are the ones where the founder has a specific, weird, earned insight that nobody else has. Not "AI for X." A genuine edge that came from living inside a problem.

The ones that are dying almost always have the same pattern: technically competent founders building something nobody asked for, moving metrics that don't matter, avoiding the conversation with the one user who'd tell them the truth.

The lucky thing is that 2nd type of founder can become the 1st kind if they don't stand still, they are willing to talk to people, try things, and always seek high rate of learning.

See 1 related tweets

@paulg: This is one of several reasons you shouldn't drop out of (or skip) college to start a startup at 18....

15. dharmesh (Group Score: 71.4 | Individual: 28.8)

Cluster: 3 tweets | Engagement: 962 (Avg: 182) | Type: Tech

RT @DavidSacks: Q: How are job postings for software engineers rising rapidly despite AI agents automating coding?

A: Because there’s far more code to manage than ever before. We’re already seeing a 14x YoY increase in GitHub commits, and it’s accelerating.

AI has dramatically lowered the cost of writing code, so it’s now being used across far more businesses, applications, and use cases.

We’re at the beginning of a massive productivity boom driven by the proliferation of bespoke software throughout the entire economy.

Coding has been AI’s breakout use case this year. The fact that it’s increased demand for software engineers — rather than decreased it — should call into question the entire “AI will cause mass job loss” narrative.

See 2 related tweets

@chrysb: this phenomenon is called Jevons Paradox:

increasing the efficiency of a resource tends to increas...

@pirroh: Some call it narrative reversal. I call it foresight.

That’s why we never stopped hiring at Replit:...

16. AlecStapp (Group Score: 69.5 | Individual: 54.1)

Cluster: 2 tweets | Engagement: 2346 (Avg: 159) | Type: Tech

The US tech industry would be a shadow of itself without immigrants.

First 10 examples that come to mind:

Elon Musk (South Africa)
Andrej Karpathy (Czechoslovakia)
Sergey Brin (Russia)
Jensen Huang (Taiwan)
Satya Nadella (India)
Ilya Sutskever (Russia)
Sundar Pichai (India)
Lisa Su (Taiwan)
Fei-Fei Li (China)
Sanjay Mehrotra (India)\n\nQT @AlecStapp: Nearly half of the founders of billion-dollar tech startups are immigrants https://t.co/oFepgmfVLE

See 1 related tweets

@GEVS94: 👀\n\nQT @AlecStapp: Nearly half of the founders of billion-dollar tech startups are immigrants https...

17. mitsuhiko (Group Score: 68.5 | Individual: 18.5)

Cluster: 4 tweets | Engagement: 527 (Avg: 555) | Type: Tech

Has been a while since I wrote about agentic engineering, so this time around some learnings of maintaining Pi as a junior maintainer to @badlogicgames :) https://t.co/TbD9Jvqk3t

See 3 related tweets

@antirez: Read this :)\n\nQT @mitsuhiko: Has been a while since I wrote about agentic engineering, so this tim...
@badlogicgames: recommended reading. i wanted to write this blog post, describing how we OSS in pi, for literal mont...
@badlogicgames: https://t.co/nWSLOTgwnE\n\nQT @mitsuhiko: Has been a while since I wrote about agentic engineering, ...

18. Shipper_now (Group Score: 67.7 | Individual: 33.9)

Cluster: 2 tweets | Engagement: 13 (Avg: 32) | Type: Tech

🚨 JUST IN: vibe-cloning is here.

copy/paste any successful business and make it yours.

this is the end of vibe coding.\n\nQT @shipper_now: 🚨 BREAKING: clone any successful company

Claude Code Opus 4.7 in @shipper_now can take any app and make it yours: design, code, business plan...

you can now ship the next duolingo / twitter / medvi / airbnb / cal ai / etc

this is THE END of vibe coding. https://t.co/3Bc84yujrN

See 1 related tweets

@chddaniel: this is f*king scary.....\n\nQT @shipper_now: 🚨 BREAKING: clone any successful company

Claude Code ...

19. alexcooldev (Group Score: 67.6 | Individual: 34.1)

Cluster: 2 tweets | Engagement: 91 (Avg: 165) | Type: Tech

A non-tech girl reached $800 MRR faster than almost anyone I’ve seen, while so many highly technical people are struggling to make any MRR.

It made me realize: if you hate distribution, find a co-founder who loves distribution. Otherwise, learn distribution yourself. AI is making the gap between developers and normal people smaller and smaller.

Anyway, distribution > building now. 🥴\n\nQT @sobedominik: my girlfriend launched her first mobile app 5 weeks ago.

today it hit $800 MRR and almost$ 1000 in revenue!

incredibly proud boyfriend moment right here.

here’s the back story:

two months ago my gf approached me and asked me if i can teach her the basics of coding.

i was really stoked because it came from her and not me.

i tried forcing a partner in the past into coding and let me tell it didn’t work out well lol

anyways, being a proud engineer and believing in teaching people first principles i decided to not throw her immediately into AI vibe coding but rather teach her the basics of coding.

we started with the Swift playgrounds course which is absolutely amazing btw

after she learned programming primitives we moved to web and i taught her a bit of typescript

we then built a very simply react native app without any AI so she could get the feeling of how apps actually work (obsly very basic but better than nothing)

at that point my life became quite busy again and i had to fully focus on work

before that though, i taught her how AI works with Cursor and Claude Code and then i went back to work

fast forward a week or two my gf approached me with her first app idea that she wants to build

i genuinely couldn’t believe it because it was actually a great idea! very rare for a first timer.

she either absorbed a lot of business things i was just casually rambling during dinners or she’s just super smart.

probably the latter.

the idea had a few rough edges but i didn’t want to influence it too much because i believe that you gotta sometimes make mistakes and learn how to fix them / pivot along the way

so i basically “left the chat” and let her cook (or fu*k up) on her own

5 weeks later and her app revenue basically reached her salary (she’s indonesia and wages are a bit rekt)

95% of her revenue is coming from organic Tiktok clips and comments

every time i see her on her phone she’s now creating new content for distribution

i can see the hunger for more in her eyes and it’s absolutely beautiful

See 1 related tweets

@seraleev: she asked to learn. she built the app. she figured out TikTok. she hit $800 MRR.

in 5 weeks.

that ...

20. rronak_ (Group Score: 67.4 | Individual: 34.6)

Cluster: 2 tweets | Engagement: 204 (Avg: 312) | Type: Tech

I’ve left Google DeepMind.

The last two years have been an incredible whirlwind.

A couple years ago, I joined a small startup called Codeium. There, I got to ship Windsurf, train SWE-1 (a frontier agentic coding model), go to DeepMind in the $2.4B acquisition. Now, I decided to leave the acquisition money and DeepMind.

I’m grateful to the mentors, teammates, and friends I worked with along the way.

At Windsurf, thanks to @_mohansolo and Douglas Chen, I got to see what a fast moving startup that ships relentlessly and builds for the future looks like. I learned from @thenickmoy how excellent research leadership can drive outsized innovation.

At DeepMind, I got to push the frontier of agentic coding, be part of the amazing team that shipped Antigravity and contributed to Gemini 3. DeepMind is a rare place: deeply curious people, exceptional research taste, and access to enormous compute and Google-scale infrastructure.

A few things that I learned:

Finding the right hill to climb. Now more than ever, there are a multitude of directions to push the frontier in AI research. It’s easy to optimize for the wrong benchmark or capability. You should step back regularly to question if you are climbing the right hill, and adjust course often.
The secret to being a fast-moving team. Moving quickly is not just about working hard and long hours. It requires making concrete bets about where the world will be in 6 months, aligning around them, and cutting everything else. This was our journey from the Codeium Extension → Windsurf IDE → SWE-1 → Antigravity → Antigravity CLI
Silicon Valley is small. Since the split of Windsurf to DeepMind and Cognition, many of my colleagues have gone to other exciting places - Thinking Machines, OpenAI, xAI, Cursor, fast-moving startups, or started their own companies. I’m grateful to have worked with so many talented, hungry people whose stories are not yet finished.

So what’s next?

We are living in one of the most exciting and powerful times in human history. Just like we transformed software engineering, soon every industry, every unit of work will be radically transformed, democratized, accelerated. With this comes new challenges, and new doors of frontier research to be opened.

More soon.

See 1 related tweets

@ns123abc: Welcome to SpaceXAI\n\nQT @rronak_: I’ve left Google DeepMind.

The last two years have been an incr...