"Andy" Antti Törrönen retweeted
Have you been sued in California for wire tapping, just for using a meta pixel?
Thousands of Ecom brands were getting shake down requests for nothing more than a conversion pixel.
Every settlement was $10,000-$50,000 bucks, right out of the pocket of small business.
I was sued multiple times.
As of today, that won’t ever happen again:
I thought this was the wrong reading of the law and wanted to do something about it.
My internal team worked with the eCommerce alliance.
They put time, money, and resources into changing this law.
I really didn’t do much.
But as a result of their hard work-
No one else will be sued for this ever again.
The work of the eCommerce alliance, backed by postscript, just saved the entire ecom industry probably 30 million dollars a year.
Huge shoutout to Gina from my team for getting this done.
Ridge could have just kept paying it.
But that’s how these things win.
Ecom brands are too fractured, everyone too cheap, everyone being crabs in a bucket.
It takes brands and great partners like postscript to work together to make the industry better.
I just saved your brand $25,000 bucks.
You can thank postscript for that.
My first OpenAI DevDay is over. If you have a chance to go, you should. It had the highest concentration of technical talent I’ve ever seen. And yes, I would probably skip the sessions next time and just learn from peers. Got a nice thing to hack, too.
"Andy" Antti Törrönen retweeted
Idag fick eleverna en målbild. De skulle få igång WordPress med MariaDB, Redis och phpMyAdmin med rootless podman och fronta allt med en reverse proxy.
Jag gav dem en checklista, "det här ska fungera när du är klar", och en server. Men inga instruktioner. Det gick fort, många var riktigt snabba.
De använde AI för att lösa uppgiften, inget fel i det. Men när jag sedan bad dem förklara varför de valt X istället för Y kunde de inte motivera. När jag frågade hur de löste två hostnames (wordpress och phpmyadmin) på samma maskin och hur de styrde trafiken rätt förstod de inte vad jag frågade. Trots att de precis löst uppgiften.
Det här är utmaningen med AI. Eleverna frågar ChatGPT och får kommandon som de blint kopierar och klistrar in i en produktionsserver utan att reflektera över riskerna. Jag måste försöka få dem att förstå att AI är ett hjälpmedel, inte lösningen. De måste iaf förstå vad de håller på med... eller?
"Andy" Antti Törrönen retweeted
Worth clarifying that this is flatly false and the two options are just two independent ways to opt-out.
We respect our users' choice whether to use their data to “improve our models for everyone” regardless of where they express that choice. Users can opt out in the in-app settings or indeed also in our privacy portal. They do not need to opt out in both places, and we will make this clearer in our Help Center.
privacy.openai.com/policies/…
There is a super shady data control setting on everyone's ChatGPT which makes it seem like your data can be used for training even if you explicitly say to *NOT* improve the model for everyone (1/9)
Readers added context they thought people might want to know
This claim is false. OpenAI has clarified that the in-app toggle and the privacy portal are independent ways to opt out of data training. You only need to use one method, and OpenAI respects the opt-out choice regardless of where it is set.
x.com/thsottiaux/sta…
help.openai.com/articles/77308…
Tekoälyalan vuoden tärkein tapahtuma, @OpenAI DevDay, on tulossa 29.9. Sain kutsun paikan päälle.
@thsottiaux vihjaa mittavasta julkaisupaketista. Veikkaukseni on GPT-6:tta hyödyntäviä sovelluksia ja toimintoja, etenkin Codexiin mutta toivottavasti myös ainakin OpenAI Prismiin.
How to get pass all the imaginary gates and evidence requirements Sol comes up with? Suddenly it needs an "independent legal review" to continue coding and evidence for 15+ other gates. How do I make it to do the work required by the gates it came up with??
Ok, the main culprit for disk use appears to be subagents duplicating a huge main thread and browser automation leaving copies in com.google.Chrome.code_sign_clone.
Total I had about 650 gigabytes of trash, 720+ gigabytes including Claude Code
@thsottiaux if helpful
"Andy" Antti Törrönen retweeted
Building a chip is expensive, long-term effort
This explains why frontier labs lost edge against Chinese labs:
By doing less.
Frontier labs are now: training/inferencing models, building harnesses, consulting/cybersecurity, building infra, designing chips & buying compute
ANTHROPIC BUILDS CUSTOM AI CHIPS FOR CLAUDE
Anthropic has confirmed it is building an in-house chip team to design custom AI processors for Claude, aiming to boost speed, efficiency and scalability.
The company will co-design hardware and AI models while continuing to rely on chips from AWS, Google, Nvidia and AMD through a multi-chip strategy.
The move follows reports of talks with Samsung as a potential manufacturing partner and mirrors a broader industry trend after OpenAI unveiled its own custom AI chip earlier this year.

















