"Cancel" comes from the Latin word for lattice. Medieval scribes voided documents by drawing crosshatched lines through them until the whole thing looked like a fence.

So technically, cancel culture has been around since the 1400s. It just used to require better penmanship.

What’s in store:

Read Time: 5 minutes

OPENAI

Yet another billion-dollar business

ChatGPT Ads has reached a $1 billion annualised revenue run rate less than 200 days after launch, according to OpenAI.

The platform now has tens of thousands of advertisers, with self-service ad buying expanding across India, Europe, the Middle East and North Africa.

Advertising is becoming a bigger part of OpenAI’s business alongside subscriptions, enterprise products and APIs.

The company says ads help support ChatGPT’s free tier, used by more than 1 billion people each week.

People already use ChatGPT to compare products, research software, look for jobs and make decisions.

OpenAI wants ads to appear during those moments when they are relevant to what users are exploring.

The system can use the current conversation to choose relevant ads.

Depending on location and user settings, it may also use wider ChatGPT activity to personalise them.

OpenAI says ads are clearly labelled, remain separate from ChatGPT’s answers and do not affect its responses. Advertisers also cannot access private conversations.

Here’s what you should know:

Your chat, now with commercial breaks

ChatGPT Ads is now available in more than 40 countries.

OpenAI also launched Ads Manager in May, giving smaller businesses a way to buy ads directly.

The platform now includes tools such as cost-per-click bidding, conversion tracking, product feeds, geographic targeting and custom audiences.

OpenAI says one ecommerce advertiser saw a 3x return on ad spend over 28 days, while one technology partner said more than 80% of ad-driven ChatGPT traffic came from new customers.

Next, OpenAI plans to expand ads into more markets and add new formats, targeting options and measurement tools.

More ads, followed by “we’re increasing the price of your subscription” - MV

Vote for live results + see results and opinions from yesterday at the bottom of the email.

AI FOR NON-TECHIES

The Cost Of A 5-Person Go-To-Market Team?

$150k+ per year.  But the cost of the AI tech stack version:  $150-300 per month. A drop in the bucket. 

Steal our Zero-Budget GTM Toolkit for build notes on five essential custom assistants that can do the heavy lifting for your business.

Spend less on…  

Scale while saving money!

Number Crunch

4 workers earn $18 per hour each. They work 7.5 hours. What's the total wage bill?

Find the answer at the bottom of the email!

That’s Weird

Japan just put 50 Qubits on the board and called it a starting point

Japan has switched on Shunkai, its first full-stack neutral-atom quantum computer. It starts with around 50 qubits and uses lasers to trap and control individual atoms, with plans to scale to roughly 500.

The bigger goal is much more ambitious: 10,000 physical qubits by 2031, plus error correction and outside access for researchers. So Japan has basically built the quantum-computing equivalent of version 1.0 and already has a very aggressive roadmap taped to the wall.

ANTHROPIC

Claude, please put the laser down

Anthropic is testing a new standard that could make it easier for AI agents to control physical machines.

The Model Hardware Standard (MHS) gives different lab and manufacturing devices a shared way to communicate with AI.

That includes microscopes, robotic arms and other specialist equipment.

Today, connecting these machines can take weeks because each one often needs custom software.

MHS could cut that setup time to hours or even minutes.

Once connected, AI agents can control several machines at once, monitor results, adjust settings and sometimes handle hardware errors without human help.

In brief:

Maybe keep a human nearby

Anthropic has already tested Claude using MHS to adjust and align a laser, while companies including AWS, QIAGEN and Universal Robots are experimenting with the system.

For now, MHS is still in testing. Anthropic says human oversight is still needed and plans more safety work before making the standard open source.

Ummmm guys? This feels like we skipped several software updates.- MG

Mindstream Picks

Health: NHS leaders warn autism and ADHD care in England risks descending into chaos as demand, costs and reliance on private providers continue to surge.

Gaming: Call of Duty: Modern Warfare 4 players have discovered a powerful new Apex Attachment that can effectively provide wallhack-style advantages, raising concerns about multiplayer balance.

Movies: Howard Shore’s The Lord of the Rings has been voted Britain’s favourite film score for the fourth consecutive year, topping Classic FM’s 2026 Movie Music Hall of Fame.

Don’t Miss: Chess.com is expanding beyond chess with Gambit, a free poker platform designed around learning and competition without real-money gambling. The company is also developing similar experiences for games like go, backgammon, and mahjong, with AI helping its small team build new platforms faster.

AI Art

Image of the Day

Daily Image Prompt

Submit your artwork to Mindstream →

Yesterday’s Poll

“Are “AI-native” workplaces the next evolution of work?”

Yes. Leaner teams and automation are the future of business. - 30%

No. It’s just corporate downsizing dressed up as innovation. - 70% ✅

Your Views:

“Even if they are successful the majority of companies will give this shift an earnest try.” - sdcoo

“if you have personal experince with doing just that you will see that potential is real, its just a bit early to be good enough. but someone will figure it out” - mark.

“I foresee the time when "Managers" will be those who manage the AI-agents who do the actual mundane work. Teams will naturally shrink down to a size reflective of this new paradigm.” - mlover

Submit your opinions in our polls to be featured!