158 Photos and videos
Pinned Tweet
Apr 19
Won 2nd place in a crazzy hackathon by @Replit
It was super fun.
I built an virality engine by reverse engineering @AIatMeta's research paper on TRIBE v2
all in all a neuro hacking for viral reels.
anyways, thanks to @amasad for giving free credits.
and @networkschool for hosting it.
If you’re in the Singapore/Malaysia area, we have a Replit Agent 4 hackathon tomorrow at Network School.
There will be a tutorial, a little remote guest appearance by my friend @amasad, and a hackathon.
Here’s the event link:
luma.com/1o8w4zlk
1
5
994
the most interesting AI shift today isn't a new model. it's that we're done picking models.
1
1
14
this is the endgame of model abundance. when you have five good-enough models, picking just one is leaving quality on the table.
1
4
my take: the 2026 meta is orchestration, not selection. routing, fusing, conducting. the conductor beats the soloist now.
3
2026 in one week:
AI companies racing to go public, and AI models getting export-controlled like weapons.
the technology grew up and got regulated at the same time.
1
5
while everyone watched the Fable 5 ban.
OpenAI quietly backed Poetic's $50M round for compliance agents, and Hypha raised $50M for AI asset intelligence.
the money is going vertical.
1
44
the open vs closed AI debate just got a third axis:
controlled. the most powerful models may end up the least accessible.
that quietly changes everything about who builds what.
2
13
Jun 13
the US government just banned the entire world from Anthropic's Fable 5 and Mythos 5.
effective immediately, on national security grounds. AI models are now controlled tech.
1
6
205
Jun 13
the US government just banned the entire world from using Anthropic's most powerful AI.
as of yesterday, Fable 5 and Mythos 5 are shut off, globally.
this is a genuinely historic moment. here's what happened.
1
63
Jun 13
the trigger, per Axios: another company claimed it jailbroke Mythos, which spooked the administration.
Anthropic says it's seen only narrow verbal evidence, is disputing it, and is working to restore access.
1
20
Jun 13
my take:
this is the moment AI models officially became national-security assets, regulated like advanced chips or weapons.
the "it's just an app" era of frontier AI is over.
12
Jun 12
everyone shows you the flashy AI agent demo.
the actual business is the unglamorous layer underneath: runtime, context, evals, governance.
that's where this week's checks went.
19
Jun 12
the chatbot era is funded and basically over.
2026's money is going two places:
the infrastructure that makes agents reliable, and robots that put AI into physical work.
plan accordingly.
16
Jun 12
the AI money this week wasn't a chatbot.
Jedify ($24M) for agent context, Niteshift ($7M, Greylock) for agent runtime, THEKER ($85M) for robots.
the boring layer is where the bets are.
23
Jun 12
THEKER just raised $85M, the largest robotics Series A in European history, with Samsung and LVMH in.
AI is leaving the screen and walking onto the factory floor.
30
Jun 12
want to know where AI is actually heading?
look at what got funded in the last 48 hours. it's not more chatbots.
it's two things: the infrastructure around agents, and robots in the physical world. here's the proof.
1
68
Jun 12
and THEKER raised $85M (Jun 11), the largest robotics Series A in European history, for factory robots that adapt without reprogramming.
the money is moving from screens to the real world.
1
22
Jun 12
the chatbot era is funded and over.
2026's money is going to the boring layer that makes agents reliable, and to the robots that put AI into physical work. that's the tell.
16