Профиль

He doesn’t come across as the most hygienic xman.

Oh goodness?! Seriously! I assumed that none of these would cause lasting damage.

Congrats!

💯 the amount of things that were no longer worth debugging now are suddenly magically solved. Also - I reboot my computer less hoping for a solution.

I was curious but didn’t end up buying the bone conduction ones after I kept on breaking my wired headphones. LMK what you think!

Wherever you want to wait two days for a query response

It’s on the broader roadmap but extremely unlikely to hit in 2026.

It is an item I'm interested in, but cost is a consideration, along with the false positive rate - since errors in these detections then slam our human moderators to unpack the inaccurate labels.

No - that's more focused on AI providers and deployers to my best understanding, and we aren't set up for that. Attie as a product isn't available in the EU and doesn't generate synthetic and manipulated content. Unless there's something else I'm missing?

This reminds me of a Charles Stross book.

So instead of addressing the root cause - they’ll instead over regulate the platforms to compensate. Ie detect real time AI enabled fraud and scams globally. *make no mistakes*

There has never seemed to be enough willpower to crack down on scaled threats across international borders and we need a new model of law enforcement or resourcing to address these cottage industries that spring up in lower resource environments.

Now being reasonable - Pandora’s box has already been opened. It’s too late since the open models today are already good enough for serious harm. Dual use can’t be put back in the box at this point. What can be addressed is the impunity that bad actors have operated with online.

Do I want better outcomes and health tracking for newborns and mothers in sub Saharan Africa? Yes. Do I want the world to be deluged in AI powered phishing via open models on the scale of email spam - no.

I used to work in international development so I think it’s great when AI is adopted by low resource govts to better aid and service their citizens. Challenge is that you get the Yahoo boys and other scammers (sometimes govt linked or even run in SE Asia) that will weaponize these at scale.

You put out op-eds and letters while furiously lobbying govt to unring said bell via regulation?

Agreed but I’d prefer to lie to myself.

My roadmaps have ended up often being aspirational due to engineering capacity…

I need to stop this. We will go bankrupt if Attie keeps on handing out free shrimp. These aren’t unlimited bread sticks.

Same!