Aadit Sheth(@aaditsh)

Basically, a Swiss Army knife can open a bottle. But a bottle opener opens a bottle way better. And ...

Aadit Sheth (@aaditsh) — Thu, 23 Jul 2026 09:54:15 GMT

Basically, a Swiss Army knife can open a bottle. But a bottle opener opens a bottle way better. And the Uber founder is now betting that factories want specialized machines over general ones.

travis kalanick@travisk

x.com/i/article/2079…

💬4🔄0❤️10👀9483📊4

Normalize beautifully written earning drops on X!

Aadit Sheth (@aaditsh) — Wed, 22 Jul 2026 22:29:48 GMT

Normalize beautifully written earning drops on X!

Sundar Pichai@sundarpichai

Q2 was an amazing quarter, with our AI investments redefining what’s possible across every part of our business.

Alphabet revenue grew 24% YoY and Google Cloud accelerated to 82% growth. We saw exciting momentum across the board from Search to YouTube to the Gemini app (which reached 950M monthly active users). Our model APIs are processing 22B tokens/min (up from 16B+ last quarter) driven by our workhorse Flash models. We’re also seeing great adoption of Gemini Enterprise, used by 90% of the Fortune 100, as well as strong demand for our security solutions.

Outstanding results and momentum, and such an exciting moment—thanks to all of our partners and employees around the world! 🙌 About to hop on the call!

💬3🔄0❤️24👀15169📊5

Nia (Narrative Intelligence Agent) is coming together very nicely. At this point, it's scouring ten...

Aadit Sheth (@aaditsh) — Wed, 22 Jul 2026 14:17:54 GMT

Nia (Narrative Intelligence Agent) is coming together very nicely.

At this point, it's scouring tens of thousands of posts a day and delivering only 3-4 on X that are actually worth our customers engaging with.

It filters out 99.9% of the BS you have to read when you're on the X timeline.

And it doesn't just match on your company name. It also catches relevant industry conversations around your product, company, or leadership. Then it surfaces what's worth engaging with and advises whether to reply, repost, or quote post.

The magic is in the filtration. There will be at least tens of posts daily that SEEM like they're worth engaging with. But they won't pass Nia's filters. Engagement isn't strong enough, or not enough influential profiles follow the original author or there might be several other reasons. The algorithm is pretty complex at this point. But what's true is simple: if Nia surfaces a post, it's likely worth engaging with. And it'll guide you on what to do.

Nia has already helped a few replies and posts hit a couple million impressions (pretty huge, when brands normally get thousands of views). I'm particularly excited to get Nia into more corporate social and exec social hands. Not everyone has the time to use X themselves. Nor do they want to. But when it comes to promoting, they still want to talk about their launch on X. That's not how it works.

X rewards you for being a PART of the conversation. Replying, quote posting, adding thoughts, extending the thread. You cannot just show up, launch your product, and get millions of views unless you spend a bomb on ads. And then you're not growing the brand organically anyway.

Nia helps you get into the right conversations on X that are trending up, at the right time, for you and your company. Excited to see how it evolves. If you'd like to check it out, DM me.

💬1🔄2❤️9👀14452📊4

The only risk is not taking any

Aadit Sheth (@aaditsh) — Wed, 22 Jul 2026 10:17:38 GMT

The only risk is not taking any

TBPN@tbpn

Whop Co-Founder & CEO @cultured’s advice to young people:

- Experiment as much as possible
- Don't feel like you're wasting time

"The only real risk is not doing anything."

Your browser does not support the video tag.

💬4🔄2❤️66👀16539📊10

Everyone's building AI agents for coding and email. Meanwhile there's a machine shop running on a w...

Aadit Sheth (@aaditsh) — Tue, 21 Jul 2026 21:57:49 GMT

Everyone's building AI agents for coding and email.

Meanwhile there's a machine shop running on a website from 2011 and a PDF catalog. Doing millions in revenue. Try buying a crane, a ship, or a custom machine part online today. You can't. Nobody's building for that.

Gushwork is building a way for buyers and sellers of the physical economy, heavy machinery, industrial equipment, things that have never had an "add to cart" button, to find each other and do business. Agentic commerce for the stuff the internet forgot.

Making the physical economy transactable is a bigger market than another coding agent. This is what they’re building toward.

Nayrhit B@NayrhitB

Today we launch the next chapter of Gushwork.

Over the last 30 years, we saw commerce moving online -

But what actually moved online were low value retail products: books, clothing, electronics.

Anything high value, from $5,000 to $1M - a hydraulic crane, an industrial roofing contract, a heat shrink machine - still runs predominantly on human buyer seller conversations.

And conversation was the one thing software couldn't do.

That just changed.

We are bringing 35 million sellers of high value products & services in the physical economy to the agentic web -

One of the largest categories of commerce that barely moved online in the internet before AI.

Your browser does not support the video tag.

💬2🔄0❤️12👀16370📊3

49% of US adults use AI chatbots (Pew Research). About 2% of households pay for one (PNC). That gap...

Aadit Sheth (@aaditsh) — Sun, 19 Jul 2026 08:18:04 GMT

49% of US adults use AI chatbots (Pew Research). About 2% of households pay for one (PNC).

That gap is wild to me. Half the country is using this stuff. Almost nobody is paying.

I think people already adopted AI. Paying for it is what's still early. And those two are getting mixed up a lot.

1993 internet was hard to get onto. But 2026 AI is free on every phone. You open ChatGPT and start talking. It's just so much easier.

My bet is that AI becomes like Google Search. The basics stay free for everyone, and get cheaper to run over time. You pay when you need the business tools (Workspace, Team etc). That's already how Claude and ChatGPT work.

Olivia Moore@omooretweets

Only 2.2% of U.S. households have a paid AI subscription

We are still so early

💬7🔄9❤️30👀19993📊10

I do my worst AI work when every message costs money. I rush the brief. Hit generate too early. Was...

Aadit Sheth (@aaditsh) — Sat, 18 Jul 2026 23:39:37 GMT

I do my worst AI work when every message costs money.

I rush the brief. Hit generate too early. Waste the credits I was trying to save.

Higgsfield's Supercomputer LLM is free now. Talk through the script until it's ready. Credits start when you generate an image or video.

Higgsfield AI 🧩@higgsfield

Free Mode is live in Higgsfield Supercomputer.

The LLM inside is now free to use for brainstorming, prompt writing, scripting, research, and project planning. Credits kick in only when you generate.

A Supercomputer-level LLM, yours to use.

Your browser does not support the video tag.

💬3🔄5❤️27👀15078📊6

Beware the man in the t-shirt not the man in the suit

Aadit Sheth (@aaditsh) — Thu, 16 Jul 2026 09:31:29 GMT

Beware the man in the t-shirt not the man in the suit

NVIDIA Newsroom@nvidianewsroom

In Tokyo today, NVIDIA CEO Jensen Huang joined leaders from Fujitsu, Kawasaki Heavy Industries, FANUC, and Yaskawa Electric. 

Together, they’re bringing physical AI from national ambition to the factory floors that power industries around the world.

💬4🔄4❤️86👀17537📊10

A great summary of what Demis is saying

Aadit Sheth (@aaditsh) — Wed, 15 Jul 2026 14:01:50 GMT

A great summary of what Demis is saying

Demis Hassabis@demishassabis

x.com/i/article/2076…

💬0🔄0❤️1👀3778📊1

I find it interesting that Sam, Satya, Sundar, Mustafa Suleyman, Chamath, Jack Dorsey... All of them...

Aadit Sheth (@aaditsh) — Tue, 14 Jul 2026 22:12:37 GMT

I find it interesting that Sam, Satya, Sundar, Mustafa Suleyman, Chamath, Jack Dorsey... All of them have endorsed this post within hours.

This almost never happens. Worth paying attention to.

Also - didn't know Demis was such a good writer :)

Demis Hassabis@demishassabis

x.com/i/article/2076…

💬0🔄2❤️1👀2820📊1

Demis posting his first X article! Very few execs doing this right now.

Aadit Sheth (@aaditsh) — Tue, 14 Jul 2026 14:59:33 GMT

Demis posting his first X article! Very few execs doing this right now.

Demis Hassabis@demishassabis

x.com/i/article/2076…

💬4🔄10❤️104👀36892📊19

I feel like this is the 107th post I've seen about people leaving somewhere exceptional and joining ...

Aadit Sheth (@aaditsh) — Mon, 13 Jul 2026 16:48:17 GMT

I feel like this is the 107th post I've seen about people leaving somewhere exceptional and joining Anthropic.

It's basically a genre at this point.

They've built something special with how they show up on X. Conversational. Relatable. Like real people.

And honestly, I think these posts are the best recruiting content on X right now. Anthropic isn't making any of them. Their people just want to share it. That's what happens when the brand is real. You don't have to ask.

Tom Blomfield@t_blom

Personal update: I'm taking a leave of absence from YC to join Anthropic. I'll be working with @NotTomBrown on the compute team.

Powerful AI has the potential to improve the life of every human on earth and, as we enter the early stages of recursive self-improvement, availability of compute becomes one of the most important issues to solve.

I'm excited to get started 🚀

💬8🔄3❤️57👀21036📊10

Jeetu, let me show you something awesome. We've built Nia (Narrative Intelligence Agent). I complet...

Aadit Sheth (@aaditsh) — Mon, 13 Jul 2026 14:25:09 GMT

Jeetu, let me show you something awesome.

We've built Nia (Narrative Intelligence Agent). I completely agree with you, the alpha on X is unreal. X will become the most important platform (if not already).

Nia goes through tens of thousands of posts daily that are specific to you as CPO at Cisco. So all the Cisco related posts. Or let's say Cisco's focus for this quarter is Cloud Control or hyperscaler infra or whatever. Then all the posts around those core narratives will be read by Nia. It will understand relevancy to you, your company, engagement growth of that post, likelihood of what it will look like in 24 hours, and after all that work, it will surface 1-2 highest leverage posts that you absolutely should engage with. It'll even tell you if you should comment, repost or quote post. I can help with that as well.

We've already seen 4x engagement on corporate and individual profiles. I'm confident you will love it. And since you are taking X more seriously now, this is the perfect time too. Just need 10-15 mins to show you how it works!

💬0🔄4❤️4👀2583📊2

Wake up babe Satya dropped his second article!

Aadit Sheth (@aaditsh) — Sun, 12 Jul 2026 16:55:39 GMT

Wake up babe Satya dropped his second article!

Satya Nadella@satyanadella

x.com/i/article/2076…

💬22🔄82❤️1525👀850796📊358

Can't believe OpenAI responded to this 95 page document with 20 words: "We have no interest in othe...

Aadit Sheth (@aaditsh) — Sat, 11 Jul 2026 16:49:26 GMT

Can't believe OpenAI responded to this 95 page document with 20 words:

"We have no interest in other companies' trade secrets. We remain focused on building innovative technology that empowers people everywhere."

NIK@ns123abc

>be Tang Tan
>24 YEARS at Apple
>VP of Product Design, iPhone AND Apple Watch
>you know every team. every project. 
>every name worth taking

>months BEFORE you leave:
>meet with OpenAI’s people
>email yourself Apple supplier intel
>apple is literally paying you while you betray them

2024: leave, co-found io with Jony Ive
2025: OpenAI buys it for $6.5 BILLION

>a one-year-old company. no product
>you’re now Chief Hardware Officer
>you’re paid in OpenAI pre-IPO shares

begin the great unbuilding of Apple

>one by one, apple’s hardware people vanish
>engineers. designers. supply chain leads

>you know which head holds which secret
>you are the mastermind 
>you pick accordingly

>interviews are not interviews
>drop secret codenames like you still work there:
> “what’s the plan?”
>candidates cram STOLEN FILES the night before like it’s finals week

>“bring Actual parts for show and tell”
>apple employees smuggling batteries and logic boards out of Apple Park in their bags
>one guy, genuinely confused: “didn’t even know we could take those from the office”
>you knew

>hand every new hire Apple’s own security manual BEFORE they resign
>the document literally lists the rules they’re about to break
>openai staff, cheerfully: “a checklist that Tang put together”

>tang did not put it together
>APPLE put it together
>tang took it on his way out

Tang Tan spent 24 years learning how Apple keeps secrets. Then 14 months teaching people how to leave with them.

APPLE IS PERSONALLY SUING HIM FOR: TRADE SECRET THEFT. BREACH OF CONTRACT. WILLFUL. EXEMPLARY DAMAGES.

💬25🔄27❤️530👀81137📊66

Zuck is back on X with a thread. Who would have thought!

Aadit Sheth (@aaditsh) — Thu, 09 Jul 2026 14:40:17 GMT

Zuck is back on X with a thread. Who would have thought!

Mark Zuckerberg@finkd

(1) Today we're releasing Muse Spark 1.1 -- a strong agentic and coding model at a very low price. It's available through our new Meta Model API and in Meta AI.

💬8🔄4❤️51👀24994📊10

This is exactly the thesis of The Narrative Company. If you're a public company CEO, X is the best ...

Aadit Sheth (@aaditsh) — Thu, 09 Jul 2026 12:29:48 GMT

This is exactly the thesis of The Narrative Company.

If you're a public company CEO, X is the best place in the world right now to control your narrative.

• The smartest engineers in the world spend time on X. If you want to hire them, you need to be active and talk about your company here
• Your market perception depends on how you show up on X. If more people see your company more often here, they'll understand what you do and where you're headed.
• It's become a baseline to have an active X. To have a thriving X, it has become a competitive advantage.

Who you should study:

@satyanadella@sundarpichai@Benioff@tobi@zoink@patrickc 

Every single public company CEO will be putting efforts in this domain. It's a no-brainer.

Look at @MicronCEO and @LipBuTan1 (Intel CEO). Both companies doing exceptionally well in the stock market and now both making active efforts to be on X and share their vision. Both accounts have the potential to grow extremely fast and take over X with great content.

All they need is great content, understanding of the X algorithm and a consistent strategy that is native to the platform. 

For example, we've been advising all our clients on writing great X articles for the last few months. It's the next big thing. Satya finally wrote his first one last month and hit 70M views. Everyone was talking about it.

Every public company CEO will have an active X in the next year. Mark my words.

Nikesh Arora@nikesharora

Have gotten a few inquiring minds - what happened to Nikesh? Why is he suddenly active on Twitter? I have to credit my friend @HarryStebbings for it. When we did the podcast he said podcasts allow you to take control of your own narrative, and so does X. "Nikesh - you should try and tweet once a day" if you have something interesting to say, your business will be perceived differently. I "rebuffed him", but then I reflected upon it and promised to give it a try, here's me trying in the last week. Should I continue?

💬4🔄4❤️24👀13814📊8

My favorite takeaway from this episode was that what used to matter was answering questions. Exams ...

Aadit Sheth (@aaditsh) — Thu, 09 Jul 2026 12:11:22 GMT

My favorite takeaway from this episode was that what used to matter was answering questions.

Exams were all about it. What do you remember from what you've learnt. Answer it in the exam.

In the AI age, it's about asking. AI can do the work. But can you ask the right questions to get what you want out of it? Can you ask things others are too scared to ask?

That's the skill now.

David Senra@davidsenra

A single, viral post on X skyrocketed Groq’s business.

Groq Founder @JonathanRoss321 was looking for a way to show people that fast inference would be extremely useful.

He remembered an example from Eric Schmidt showing off an LLM at a conference three weeks before the ChatGPT moment. Nobody in the audience cared.

Then he looked at what made ChatGPT’s launch successful and realized that it worked because they allowed people to ask their own questions.

“They got an answer to their question that was specific to them. That was magical.”

“Seeing text show up for someone else's answer wasn't magical.”

So he put Groq’s fast inference hardware on the internet for anyone to try.

Someone posted a video on X of an LLM running on Groq running super fast.

“It was just such eye candy when people saw it that everyone started creating their own and we just went viral.”

Your browser does not support the video tag.

💬5🔄1❤️25👀11710📊7

OpenAI just launched GPT-Live. Full-duplex voice, delegates hard questions to GPT-5.5 in the backgro...

Aadit Sheth (@aaditsh) — Wed, 08 Jul 2026 22:23:27 GMT

OpenAI just launched GPT-Live. Full-duplex voice, delegates hard questions to GPT-5.5 in the background. The intelligence jump is real (0.7% to 75.2% on their own web research benchmark).

I recently got a call from a restaurant confirming my reservation. It was AI. I could tell pretty quickly. Once I realized, I started asking it questions to test it. It just hung up on me. Not a great experience really.

So I'm genuinely excited about voice AI getting better. It only works when it feels completely human. Even 98% is off. You just feel it.

OpenAI@OpenAI

Introducing GPT-Live, a new generation of voice models for natural human-AI interaction.

Rolling out in ChatGPT starting today.

You’ll want to turn the sound on for this one.

Your browser does not support the video tag.

💬8🔄4❤️48👀22041📊11

I think this is close but incomplete. It's more like a sandwich. The human decides what's worth mak...

Aadit Sheth (@aaditsh) — Wed, 08 Jul 2026 15:28:51 GMT

I think this is close but incomplete.

It's more like a sandwich. The human decides what's worth making. AI does a lot of the middle stuff. The stuff that used to take hours but not much human thought.

Research, drafting, processing. Then the human finishes with taste.

The part Jack doesn't say is why humans have to finish it. AI is too agreeable. If you think something is good, AI will tell you it's good. If you think it's bad, it will agree it's bad. It just mirrors you. If you leave it to AI, it will say it's ready when it clearly isn't.

A human will internally feel whether something is right. If something is trash, you'll know "this is bad, let's just start over."

AI doesn't have that.

So it's really human-fed, AI-middle, human-finished.

jack@jack

ai-fed, human-finished.

💬3🔄3❤️26👀10681📊5

I can tell if something is AI-generated within 10 seconds. Just like people who vibe code a lot can...

Aadit Sheth (@aaditsh) — Tue, 07 Jul 2026 22:19:17 GMT

I can tell if something is AI-generated within 10 seconds.

Just like people who vibe code a lot can tell if a website was vibecoded. They know what fonts Claude usually picks. What the boxes look like. The formatting, etc. They've just seen it enough times to know that it just is. It's the same thing for me with writing. Seen it enough times.

That's not to say that you shouldn't use AI for writing. I use AI constantly. For things like research, checking facts, interviewing me with thoughtful questions or even testing my arguments. It's good at these things. But the ideas and opinions have to come from you. A lot of the times my ideas come from conversations or something I read or just a life experience. AI cannot do that.

I think a lot of people started writing post 2023 (ChatGPT launch). And honestly now it's just sort of assumed that AI is used to a certain extent in all writing. That's fine. The problem comes when the ideas are obviously not yours. It's just seen as lazy and not well done.

This matters way more the bigger you are. If you're just starting out, nobody cares enough. But a big CEO posting AI content will be seen as lazy, unprepared and a try hard.

💬13🔄2❤️40👀12155📊16

This is being framed as an image model but it's actually @Meta's ad business in a new costume. 98% ...

Aadit Sheth (@aaditsh) — Tue, 07 Jul 2026 21:24:36 GMT

This is being framed as an image model but it's actually @Meta's ad business in a new costume.

98% of Meta's Q1 revenue ($56.3B) comes from ads.

Advantage+ is already cutting cost per lead for advertisers. And Muse Image goes into Advantage+ within weeks.

@finkd basically told investors the goal is businesses just tell Meta what they want and how much they'll pay per result. Muse is what makes that work. You generate the creative for basically nothing. Test a thousand variants. Let the algorithm pick the winner.

I think the part that should worry creative agencies is pretty obvious. If generating and testing ad images costs close to zero, what happens to the agencies charging $50K per campaign to do the same thing?

Meta invested $14.3B in Scale AI and brought @alexandr_wang in to lead their new lab. The first products from that lab are a consumer image tool and an ad pipeline. The ad pipeline is the one that pays for everything.

Alexandr Wang@alexandr_wang

1/ releasing muse image today — the first image generation model from MSL. it's agentic: pairs with muse spark to reason through your prompt, search the web, and plan before it generates. people get what they meant on the first try. live now in the Meta AI app.

💬8🔄4❤️23👀11360📊9

China gave its models away when it was behind. Qwen, DeepSeek, GLM. All were open. The strategy was...

Aadit Sheth (@aaditsh) — Tue, 07 Jul 2026 19:04:18 GMT

China gave its models away when it was behind.

Qwen, DeepSeek, GLM. All were open. The strategy was to commoditize the model layer so US labs can't charge premium prices.

It worked. US companies now route over 30% of their AI tokens through Chinese models every week (a year ago that was 11%).

Now Chinese models are roughly 6 to 9 months behind the US frontier instead of years. And this week it's reported that Beijing is discussing restricting overseas access to its best models for the first time.

The logic is pretty simple. You give the model away when you're second. You restrict it when you think you're about to be first.

Ethan Mollick@emollick

This is a key reason I don’t expect the flow of frontier open weights models to continue indefinitely, or even for very much longer. reuters.com/world/beijing-…

💬53🔄36❤️293👀64052📊75

Here's the link to apply: https://t.co/cXUHpxU2E7

Aadit Sheth (@aaditsh) — Fri, 03 Jul 2026 17:34:38 GMT

Here's the link to apply: thenarrative.company/careers

💬0🔄0❤️2👀2791📊1

We're hiring a Narrative Engineer. Here's what the job actually is: you'll be writing for and strat...

Aadit Sheth (@aaditsh) — Fri, 03 Jul 2026 17:34:35 GMT

We're hiring a Narrative Engineer.

Here's what the job actually is: you'll be writing for and strategizing with the comms team of multi-hundred billion dollar companies: enterprise SaaS, semiconductors, AI etc.

Before anything goes on the CEO's X or LinkedIn, it goes through you.

Some of what that's looked like this past year:

→ We delivered the highest-performing post a Fortune 500 CEO had ever put out. He'd been posting for five years.

→ We built software that scans 10,000 X posts a day, surfaces the 2-3 best opportunities to engage, and the result outperformed the client's average by 25x.

You'll shape what the most important companies in the world say before they say it. You'll understand how the biggest leaders think about narrative at the moments that matter most.

Honestly, I've learned more about how the world actually works in the past year than in the five years before it. It's truly an n of 1 position.

DM me or apply below.

💬12🔄7❤️74👀14008📊29

Yay Fable 5 is back! But the bigger signal is what Anthropic shipped alongside it: Sonnet 5 at $2 pe...

Aadit Sheth (@aaditsh) — Wed, 01 Jul 2026 12:36:55 GMT

Yay Fable 5 is back! But the bigger signal is what Anthropic shipped alongside it: Sonnet 5 at $2 per million tokens (less than half the price of GPT 5.5).

Developers moved to other models during the 18 day ban. Of course not everyone will come back just because fable is online again, but a brand new model at a fraction of the cost will bring a lot of them back.

Anthropic@AnthropicAI

We’ve received notice that the Department of Commerce has lifted export controls on Claude Fable 5 and Mythos 5.

We'll begin restoring access tomorrow, and will share an update soon.

We’re grateful to our users for their patience, and to everyone who worked with us on redeploying the models.

💬6🔄5❤️40👀19455📊8

Two years ago the AI debate was about which model is smartest. Now Cursor shipped a mobile app so yo...

Aadit Sheth (@aaditsh) — Mon, 29 Jun 2026 21:22:05 GMT

Two years ago the AI debate was about which model is smartest. Now Cursor shipped a mobile app so you can build software while you're walking your dog.

I think the best AI in the world doesn't matter if you only use it at your desk 8 hours a day. The one that's in your pocket when you have an idea at midnight is the one you'll never stop using :)

Cursor@cursor_ai

Introducing Cursor for iOS.

Build from anywhere by launching always-on cloud agents. Or remotely control agents running on your computer from the app.

Composer 2.5 is 75% off in the app now through July 5.

Your browser does not support the video tag.

💬6🔄2❤️17👀11037📊7

SpaceX went public on June 12, bought Cursor for $60 billion on June 16, and announced Grok is train...

Aadit Sheth (@aaditsh) — Sun, 28 Jun 2026 13:39:03 GMT

SpaceX went public on June 12, bought Cursor for $60 billion on June 16, and announced Grok is training on Cursor data today.

7 million developers use Cursor every day and most of them started using it because it ran on Claude. Now the code they write is training Grok instead.

I think the benchmark comparison is going to get all the attention but this is the most interesting part of the whole announcement and nobody is really talking about it.

Elon Musk@elonmusk

Grok 4.5, based on our 1.5T V9 foundation model, with Cursor data added in supplemental training, is now in private beta at SpaceX & Tesla. Early evals show performance close to, perhaps exceeding Opus. 

RL is continuing to significantly improve the model, and the Grok Build harness gets better every day.

Nice work by all those involved!

Completely trained from scratch new models will be released by @SpaceX every month this year.

💬39🔄58❤️938👀133033📊110

AI labs say their technology will cure cancer in a few years. But when Chinese companies ran million...

Aadit Sheth (@aaditsh) — Fri, 26 Jun 2026 20:39:43 GMT

AI labs say their technology will cure cancer in a few years. But when Chinese companies ran millions of fake queries to copy their AI over six weeks, a human security team caught it afterwards. The AI didn't catch it at all. I think Gurley is right to ask the question honestly! If we're this close to superhuman AI, why does it still need humans and politicians to protect it?

Bill Gurley@bgurley

If you are on the verge of AGI or ASI, why isn’t your model smart enough to recognize espionage distillation in real time? You say “cure cancer in a few years.”  Isn’t sniffing illicit distillation quite a bit easier than curing cancer? Why write letters to DC? Just use AGI.

💬1🔄0❤️14👀9330📊2

Your MacBook Air just got $200 more expensive and it has nothing to do with Apple. AI companies are...

Aadit Sheth (@aaditsh) — Thu, 25 Jun 2026 21:15:01 GMT

Your MacBook Air just got $200 more expensive and it has nothing to do with Apple.

AI companies are buying so much memory that chip prices are spiking for everyone (Apple, Asus, Dell, Alibaba etc.) are all raising prices at the same time.

 You're basically paying an AI tax on every device you buy now whether you use AI or not and I don't think this goes away anytime soon.

Marques Brownlee@MKBHD

Memory price hikes have finally started hitting Apple. Today, they raised some base prices:

Macbook Neo: $599 -> $699
Macbook Air: $1099 -> $1299
Macbook Pro: $1699 -> $1999
Mac Studio: $1999 -> $2499
iPad Air $599 - $749
iPad Pro: $999 -> $1199

💬36🔄223❤️1801👀168730📊211

First time experiencing @Starlink on @emirates. It’s magical. Faster wifi than most offices, easy t...

Aadit Sheth (@aaditsh) — Thu, 25 Jun 2026 06:37:45 GMT

First time experiencing @Starlink on @emirates. It’s magical.

Faster wifi than most offices, easy to connect (used to be a pain) and completely free (used to be 20 usd).

Well done @elonmusk and team!

💬4🔄0❤️11👀5076📊5

Honestly the most important line here: "It wasn't a better model. It was CONTEXT." Everyone is deba...

Aadit Sheth (@aaditsh) — Tue, 23 Jun 2026 21:16:59 GMT

Honestly the most important line here: "It wasn't a better model. It was CONTEXT."

Everyone is debating GPT vs Claude vs Gemini. It doesn't matter. AI that actually knows your work will beat a smarter model that doesn't. Every single time.

I see this with every company we work with. They use the same tools but get completely different results depending on how much context the AI has. Most people are still copy pasting into a blank ChatGPT window and that's okay for now until they figure out what's actually possible in 2026.

Zeb Evans@DJ_CURFEW

The 100x org went viral. Half the internet hated it. The other half was curious. 
 
One month later: output is up. productivity is spiking. we're approaching a 5:1 agent-to-human ratio.
 
And contrary to popular belief, we're doing the OPPOSITE of tokenmaxxing. We're tokenSAVING....
 
Why 5,000 agents for 1,000 people? Because agents AREN'T workflows... Agents define permissions and skills. the vast majority of ours don't bother a single human, they run as triggers and loops in the background.
 
My rule is simple: if you're doing work that ISN'T purely using your judgment → build an agent or a system to remove that bottleneck. 
 
So that's what we did. And yes, it broke things. Honestly it created real pain for first couple weeks. To be  expected.
 
we have TONS of examples now people who had NEVER built a system in their life and are now running large-scale agentic systems that power much of our company.
 
I always say - if you automate your own job with AI, you'll ALWAYS have a job. Systems is arguably the most important thing of all.
 
Today, regardless of what your role used to be, everyone does the IC work end-to-end:
→ Before, you riffed on messaging. Now you ship the landing page yourself.
→ Before, you were a PM waiting on design and research. Now you're a builder iterating in real-time.
→ Before, you waited on someone to finish something. Now you don't.
 
The 100x future isn't coming. It's HERE. Bottlenecks are mostly gone and those that aren't are in sight. 
 
And obviously gotta lead by example. I even designed and built the Brain² landing page. I built most of the product. The in-app announcements. The emails. All of it.

Breaking our existing processes let us rethink everything WITHOUT prior constraints. 
 
Enter, Process Miner: my agent that mines all the processes in our company. This is just one of the infinite use cases for setting up an LLM to process EVERY SINGLE event that happens in your company. 
 
These types of agents simply are not sustainable without compressed context. Instead of starting from scratch each time and deep researching across the ~100,000 daily activity items in our company, it starts from context that has already been processed, summarized, and organized. 
 
Everyone's jumping on the context and memory train now. I love it. We've been building this for two years. and we STARTED OVER 4 times! We came out the other side with what we call live intelligence:
→ LLM pipeline: every event runs through a cheap LLM that summarizes, organizes, triggers, and rolls up

→ Self-organizing memory. Don't overcomplicate it. Start simple. 

→ Self-improving orchestration. THIS IS THE KEY. BUT... don't over-focus on orchestrating models... 
it's WAY more about orchestrating CONTEXT. 
 
It wasn't more tokens. It wasn't a better model. It was CONTEXT.
 
Every LLM is average by default. Think about that. If you're using the same model as everyone else, your intelligence is by definition, AVERAGE.
 
That's what really changed everything for us internally and it's why we even have a product to release in the first place.
 
Our company AI is personalized (and yes these learnings are applicable regardless what system you use). 

Every time you use Brain², we dynamically rewrite the system prompt based on real-time context and intent. Your activity, your decisions, your projects, your memory, your preferences. 
And your FEEDBACK.. that's WAY more important than decisions that everyone goes overboard on being so valuable. 
All injected AHEAD of time. 
 
When AI has 100% context, you don't need prompts in the same way. You need INTENT.
 
I go anywhere in ClickUp and type "@brain slides." Done. "@brain campaign." Shipped. "@brain app." Live.
 
And it REMEMBERS. Every interaction compounds. One Brain for your whole company, and every person makes it smarter for everyone else.
 
I flew our AI team to my house for a week to finish this. Monitors and whiteboards everywhere. SO proud of this 100x team. 
 
Brain² is truly the best work AI in the world... and I firmly believe it's also the best personal AI. Naturally, easy for me to say and honestly I assumed it was just my bias originally. 
but then EVERYONE who tried it started saying the same thing. 
 
So we tested it. A three-week study, randomly recruited participants, blind pairwise comparisons. They picked Brain² as the best nearly 100% of the time over ChatGPT, Gemini, and Claude.
 
Here's the kicker: you can run ANY of those models inside Brain². Same frontier models. Radically different results. 

Brain² doesn't win because the model is better. It wins because of CONTEXT. It learns from every single interaction and new piece of context it sees. 
 
OH and we built our harness around being OBJECTIVE. Soo it can be sassy - you will definitely see some 'fun' answers in your day to day. 

AI that is built to agree with you is NOT productive. That makes no one better. 
 
Run it against whatever you're using right now. It compounds in quality the more that you use it. 

Prove me wrong → clickup.com/brain

Brain² is your company's AI. 

I built this for us originally to realize 100x, and now we're shipping it externally. Even if you're building your own I just hope sharing in public helps in some small way.
 
💜 Let's make the world more productive

Your browser does not support the video tag.

💬11🔄1❤️16👀11383📊10

Meta paid $900M for a 20% stake in CRED (no board seat and no access to customer data). Basically a...

Aadit Sheth (@aaditsh) — Mon, 22 Jun 2026 18:41:29 GMT

Meta paid $900M for a 20% stake in CRED (no board seat and no access to customer data).

Basically a $900M hire. The Cred stake is the wrapper.

Kunal built the company from $1m of his own capital to $300M+ in annual revenue and around 20M members.

First profitable quarter this year. IPO was next.

He's leaving all of it to run WhatsApp and its 3 billion users as a full time employee.

Kunal Shah@kunalb11

It’s been a minute. 

2015–2018
- Exited FreeCharge. Spent time learning and investing.
- Pondered about: Why can't trust be rewarded? Started with $1M of personal capital.
- Launched CRED to reward people for paying credit card bills on time.

2019–2025
- Built a system run by a team that values ownership, judgment, and craft.
- Grew from 0 to 17M members by aligning incentives with behaviour.
- Built several products during COVID lockdowns.
- Raised $900M+ from global investors. Did 4 ESOP buybacks.
- Made Indiranagar and IPL ads slightly more interesting. 
- Received a full stack of regulatory licences.
- Lost 35 kilos.
- Scaled from 0 to ~$325M ( ~₹3,200 crore) in annual revenue across payments, lending, insurance, commerce, wealth, and credit cards.
2026
- First profitable quarter (yet occasionally asked what our business model is)
- Raised another $900M from Meta in primary and secondary capital.
- Announcing our 5th ESOP buyback.

Today

CRED is ready for its next phase. I am stepping back and @miten steps in as interim CEO, partnered with an incredibly talented team. He has been heading strategy and finance and suffering me since 2020. I’m stepping away from the operating role and will continue as a shareholder. My commitment doesn’t change. Just the role. 

Extremely grateful to our members, partners, regulators, and investors who made this possible. And to our board, Shailendra, Micky, Saurabh for their extraordinary conviction. 

Team CRED, I’ll still expect you to be a 10x version of yourselves.

As for me, I’ll be joining Meta to lead WhatsApp globally.

Meta comes in as a minority investor in CRED. No access to member data. 

While it’s come very far, the delta between WhatsApp today and its full potential is massive. I look forward to working with Mark, Chris, and the leadership across Meta for the next step in WhatsApp’s journey. Will, thank you for scaling something the world relies on quietly, and for making this transition smooth.

Onwards.

💬12🔄8❤️134👀52456📊23

What it takes to join OpenAI: 57 interviews 46 recruiter calls 16 post offer chats And she still sa...

Aadit Sheth (@aaditsh) — Mon, 22 Jun 2026 05:29:31 GMT

What it takes to join OpenAI: 57 interviews 46 recruiter calls 16 post offer chats

And she still says the biggest factor was having someone inside the company vouch for you.

One referral is worth about 40 cold applications in AI research hiring right now. The formal process is the backup channel. The network is the actual channel to get interviews!

Alisa Liu@alisawuffles

I'm joining OpenAI next week!🥹 The job search turned out to be really challenging but also super rewarding, so I wrote a small blog to share what I learned along the way and hopefully make the process a little less mysterious for the next person.alisawuffles.github.io/blog/job-searchD

💬21🔄39❤️790👀204651📊132

Anthropic is going big on biology. In April they paid $400M for a biotech team. Now they hired the N...

Aadit Sheth (@aaditsh) — Sat, 20 Jun 2026 14:31:35 GMT

Anthropic is going big on biology. In April they paid $400M for a biotech team. Now they hired the Nobel prize winner who built AlphaFold.

In 2024, @DarioAmodei wrote that AI could compress 100 years of biological progress into 10. He definitely meant it.

I really think biology is the most important use case in AI right now. I don't know much about it but I think it's pretty incredible what humans have managed to do with drug discovery.

Now imagine humans + AI at that same problem. Things just accelerate! Helping people live longer and better.

John Jumper@JohnJumperSci

A bit of news: After nearly 9 years, I have decided to leave Google DeepMind and join Anthropic (after taking some time to recharge). I am incredibly grateful for my time at GDM. @demishassabis took a real chance letting me lead the AlphaFold team just six months after finishing my PhD, and the entire GDM team taught me so much about how to do great science. GDM is a special place, and I’ll still be excited to hear about what amazing things they discover next.

💬20🔄53❤️520👀60524📊77

Google paid $2.7 billion in 2024 for a deal structured primarily to bring Noam back from Character(....

Aadit Sheth (@aaditsh) — Thu, 18 Jun 2026 10:07:07 GMT

Google paid $2.7 billion in 2024 for a deal structured primarily to bring Noam back from Character(.ai)

He left less than two years later for OpenAI.

That's over $100 million per month. The person who co-authored the famous Transformer paper and reportedly "saved Gemini" with a few lines of training code is now building architecture at OpenAI.

This is what the AI talent market looks like in 2026. Everyone is pushing the limits at the frontier!

Noam Shazeer@NoamShazeer

I’m excited to share that I’ll be joining OpenAI and look forward to working with the exceptional team there.

It was a difficult decision to move on. I’m incredibly proud of the amazing team at Google and everything we’ve built together.  It has been an honor and a pleasure to work with all of you.

💬15🔄9❤️187👀42736📊30

For context: Amazon's CEO flagged Claude Fable 5 to the White House five days ago. It got banned fro...

Aadit Sheth (@aaditsh) — Wed, 17 Jun 2026 19:22:57 GMT

For context: Amazon's CEO flagged Claude Fable 5 to the White House five days ago. It got banned from foreign access within 24 hours.

Amazon is also Anthropic's largest investor at $4B.

Dario's models are banned. Sam's aren't.

Chubby♨️@kimmonismus

Trump and the AI ​​CEOs are currently meeting. Judging by his facial expression, Sam seems to be enjoying himself more than Dario.

Your browser does not support the video tag.

💬5🔄11❤️106👀25730📊16

Up until this point, I've been pretty skeptical of vibecoding as a means of creating anything of sig...

Aadit Sheth (@aaditsh) — Wed, 17 Jun 2026 16:17:43 GMT

Up until this point, I've been pretty skeptical of vibecoding as a means of creating anything of significant enterprise value. 

I’m a believer now. Watch this and drop your $1M business below 👇

Sid Yadav@sidyadav

1. Circle AI

An AI partner that builds, runs, and grows your digital business with you.

Describe your dream business and bring it to life with Circle AI.

Your browser does not support the video tag.

💬2🔄3❤️13👀4921📊4

Best merch I’ve seen: SpaceX Cursor OpenAI Anthropic Google

Aadit Sheth (@aaditsh) — Wed, 17 Jun 2026 15:41:10 GMT

Best merch I’ve seen:

SpaceX
Cursor
OpenAI
Anthropic
Google

💬1🔄0❤️8👀1871📊2

Cursor's developer market share dropped from 41% to 26% in the last twelve months. And revenue doubl...

Aadit Sheth (@aaditsh) — Tue, 16 Jun 2026 20:12:17 GMT

Cursor's developer market share dropped from 41% to 26% in the last twelve months. And revenue doubled to $4B ARR in the same period.

Both are true at the same time.

The $4B is real but it's mostly enterprise contracts that Fortune 500 companies signed months ago. And now  developers are moving to Claude Code and GitHub Copilot.

I think this is a great outcome for the founders btw. Built something incredible in just four years. But the question for SpaceX is whether they just paid $60B for a business whose best growth quarter is already behind it, or whether the enterprise contracts keep compounding.
 
I don't know but I believe never bet against Elon. But I'm excited to see how this plays out!

SpaceX@SpaceX

SpaceX has exercised the option to acquire @cursor_ai in an all-stock transaction with the goal of building the world’s most useful AI models.

For the past few months, SpaceXAI has been jointly training a model with Cursor, which will be released in Cursor and Grok Build soon.

We look forward to working closely with the Cursor team to advance our frontier AI capabilities

💬43🔄3❤️214👀57741📊48

Jensen spent 30 years building NVIDIA before anybody understood what GPUs would become

Aadit Sheth (@aaditsh) — Tue, 16 Jun 2026 06:37:10 GMT

Jensen spent 30 years building NVIDIA before anybody understood what GPUs would become

Paul Graham@paulg

I was talking to a YC partner about how well all the hard tech startups are doing. He said investors are hot to fund them because they're afraid AI will eat all software. I'm glad hardware startups are getting funded, but this is a mistake. Good founders are what wins.

💬3🔄2❤️29👀11550📊5

Australia tried the exact same ban. It's not going to work. They removed like 5M minor accounts in ...

Aadit Sheth (@aaditsh) — Mon, 15 Jun 2026 11:38:58 GMT

Australia tried the exact same ban. It's not going to work.

They removed like 5M minor accounts in five weeks and 70% of kids were still on the apps. I think the intent is good but honestly I don't see how this works.

Keir Starmer@Keir_Starmer

We are banning social media access for under 16s.

These days kids must find their feet in a world where technology intrudes into every area of their life.

I just can’t let that go on anymore. So we’re giving children their childhoods back.

Your browser does not support the video tag.

💬6🔄0❤️6👀2458📊5

Satya posting his first X article! Very few execs doing this right now.

Aadit Sheth (@aaditsh) — Sun, 14 Jun 2026 17:42:16 GMT

Satya posting his first X article! Very few execs doing this right now.

Satya Nadella@satyanadella

x.com/i/article/2065…

💬7🔄9❤️143👀88630📊24

It still blows my mind that Andrej Karpathy (who led Tesla’s Autopilot AI) dropped a 3.5-hour free c...

Aadit Sheth (@aaditsh) — Sat, 13 Jun 2026 16:53:43 GMT

It still blows my mind that Andrej Karpathy (who led Tesla’s Autopilot AI) dropped a 3.5-hour free course on how ChatGPT actually works.

In a world obsessed with selling knowledge, he just gives it away. Save this one.

💬9🔄70❤️347👀16828📊111

Anthropic is projecting its first profitable quarter ever ($559M). SemiAnalysis says Claude Max del...

Aadit Sheth (@aaditsh) — Sat, 13 Jun 2026 08:32:14 GMT

Anthropic is projecting its first profitable quarter ever ($559M).

SemiAnalysis says Claude Max delivers $8,000/mo in compute for $200.

Something doesn't add up, right?

SemiAnalysis ran the math on Claude Max and ChatGPT Pro. Bought each plan, maxed out the usage, and calculated compute value at API rates. 

$8,000/mo for Claude. $14,000/mo for ChatGPT. Everyone concluded the same thing: these companies are burning cash to buy your loyalty. You probably knew that already.

Interestingly, planet Fitness is profitable because 60% of members don't show up in a given month. They price low enough that non-users don't bother canceling.

The $200/mo AI plans work the same way.

The $8,000 figure assumes you hit the cap. Most subscribers don't. The $200 tier is basically for people who've hit a specific wall. Most people haven't.

So the real math isn't "$200 for $8,000 in compute." It's: what does the median subscriber actually use? If it's 10-15% of the cap, the blended economics look completely different. The heavy users aren't subsidized by Anthropic's balance sheet. They're subsidized by the subscribers who pay $200 and barely touch it. This was me btw. Now back on the $100/mo plan.

This is the business model. Airlines, insurance companies, and every gym on earth run this way.

There's one number that would settle this: the actual usage distribution across Max subscribers. What percentage hit 90% of the cap? What percentage stay under 20%? Nobody has published it. Not SemiAnalysis, not the financial press, not Anthropic.

I'd love to see this chart but my guess is that there are a lot of people that are using it less than 50% of the limit.

Chubby♨️@kimmonismus

Subscription plans are massively subsidized.

And by massively, I mean absurdly:

Claude Max 20x: $200/month, with usage reportedly worth around $8,000

ChatGPT Pro 20x: $200/month, with usage reportedly worth around $14,000

💬11🔄6❤️26👀13033📊13

Cognizant, Deloitte, KPMG, Accenture, and TCS all signed Claude partnerships this year. That's over...

Aadit Sheth (@aaditsh) — Thu, 11 Jun 2026 11:26:06 GMT

Cognizant, Deloitte, KPMG, Accenture, and TCS all signed Claude partnerships this year.

That's over a million employees getting access to the same AI model.

But dug a little deeper to find a different story underneath...

Cognizant gave Claude to 350,000 employees (basically their entire company). Stock is down 36% this year.

Accenture gave it to 30,000 (3.8% of their workforce). They booked $2.2 billion in AI work last quarter. Nearly doubled YoY.

TCS gave it to 50,000 (8.3% of their workforce). Their AI services line is at $1.8B annualized, growing 17.3% QoQ.

The two firms with the smallest rollouts have the biggest AI revenue. The firm that gave access to everyone hasn't disclosed comparable numbers.

I think this tells you something about how AI actually gets adopted inside large companies. Turning it on for everyone is easy. Figuring out where it creates revenue and deploying there first is hard.

Accenture and TCS picked specific functions: engineering, finance, legal, sales. My guess is they adopted it better.

This is the AWS pattern again. When cloud arrived, the companies that won weren't the ones that migrated everything overnight. They were the ones that figured out which workloads to move first.

Anthropic is still the platform underneath all of them. Every firm competes with each other. Anthropic collects from all of them.

But honestly, I think the real question for anyone evaluating these firms isn't whether they use AI. It's whether they know where to point it.

That's the hard part. Always has been with technology.

💬10🔄7❤️47👀11402📊18

Aadit Sheth (@aaditsh) — Wed, 10 Jun 2026 05:13:40 GMT

💬1🔄5❤️130👀66897📊29

1. Applied Mathematics 2. Bioengineering

Aadit Sheth (@aaditsh) — Wed, 10 Jun 2026 05:13:37 GMT

1. Applied Mathematics
2. Bioengineering

💬4🔄14❤️205👀77639📊49