Joined October 2021
- Tweets 120
- Following 100
- Followers 48
- Likes 172
25 Photos and videos
Jun 13
Hahaha, human pride strikes again 🤷♂️
Jun 13
The US government, citing national security authorities, has issued an export control directive to suspend all access to Fable 5 and Mythos 5 by any foreign national, whether inside or outside the United States, including foreign national Anthropic employees.
The net effect of this order is that we must abruptly disable Fable 5 and Mythos 5 for all our customers to ensure compliance.
Access to all other Claude models is not affected.
We apologize for this disruption to our customers. We believe this is a misunderstanding and are working to restore access as soon as possible.
Read our full statement: anthropic.com/news/fable-myt…
10
Roberto Lucas retweeted
Jun 10
Teaching TCP/IP this way would save thousands of students.
40
693
5,145
197,016
Jun 12
I've been running a plan that shows how #SoftwareEngineering products are built today. Watching #AI mature gave me insight to understand how they work and, naturally, build a team of agent with roles, skills, tools and responsibilities using same the protocols.
🧵
1
1
33
Jun 12
Those questions need a human in the loop. Without clarity on what you're doing, and without governance over what your agents are capable of, you aren't using your #AI properly. Governing the AI is the point, same as its security and reliability.
1
20
Jun 12
I am refreshing my site to reflect what we do at Andler Develops, a #FrontierEngineer shop that ships Web3 AI crossover systems. Content, articles, and where I'm doing things live, all going up there. Stay tuned!
26
The institutions that govern high-risk technology are built before the crisis, not after it.
The IAEA wasn't built after a nuclear catastrophe. It was built in the window when the stakes were understood but the disaster hadn't arrived yet.
That window is open for AI governance today.
The institutions built in this window define the terms. The ones built after inherit the compromises.
1
33
May 1
Google Cloud: "Never Expire sessions? Over. 16 hours max." #CyberSecurity
This isn't bureaucracy. It's a scream into the darkness.
April 2026: X faced a botnet that compromised 4.8M accounts. 18 workers. Every root password exposed.
Something broke. And we know it. 🧵
1
49
May 1
Traditional security assumes human attackers:
❌ Slow
❌ Need sleep
❌ Cognitive limits
❌ Manual scale
AI: #AIGovernance
✅ 24/7 w/o pause
✅ Global scale in seconds
✅ Thousands of simultaneous vectors
✅ No fatigue errors
The model is broken. And how many say it out loud?
1
27
May 1
Google, X, Microsoft: all shortening exposure windows. All reacting to the same threat. #AIRisk
The question is: Can we coordinate institutional defenses faster than AI automates offenses?
If not, it's not just digital integrity that's compromised. It's human.
18
Apr 15
I was reading one of the dreams that Wobblus is having now that @openclaw has had this feature for a few days now. This, is impressive and you can tell the difference of continuity between sessions. If you keep syncing their memory and indexing, all good 😉
#AgenticAI #AI
35
Mar 11
Just watched "AI Is A Massive Problem. Here's Why." It crystallizes what many of us in #AISafety have been saying.
The issue isn't AI capability. It's #AIGovernance. Let me share why the current trajectory is not stable, and what we need to do about it 👀
#ExistentialRisk
1
1
42
Mar 11
No need for autonomous general AI to solve all problems. Safer paths exist. But they require restraint.
The question: Will we build governance infra before crisis, or after?
History suggests "after" rarely works.
youtu.be/A3HjNYDIhGU?si=uMoF…
#ResponsibleAI #AIFuture #TechEthics
1
68
Mar 11
PD: I recommend watching the #WestWorld series (not the last season), it is a sci-fi that has an autonomous general AI and where AI and humans are seamlessly together... 😁
25