Jack (@itsjack) header
Jack (@itsjack) profile picture

Founding engineer @schematikio | Building agents that build hardware | ex-Head of Product & Tech @thenextweb, @Vinted

Amsterdam, The Netherlands
Joined August 2019
Photo by Jack (@itsjack)Photo by Jack (@itsjack)Photo by Jack (@itsjack)Photo by Jack (@itsjack)Photo by Jack (@itsjack)Photo by Jack (@itsjack)Photo by Jack (@itsjack)Photo by Jack (@itsjack)Photo by Jack (@itsjack)Photo by Jack (@itsjack)
solid reference
incredible things happening on techtok
222
Migrating from @conductor_build? Here's what you should do
I am very excited to share that Conductor is joining Meta! We have some big, crazy plans for the next interface. If you liked Conductor, you’re gonna love what’s coming. We’re super excited to work with @natfriedman and @alexandr_wang at MSL to build it. They’ve been on an absolute tear, and I’m convinced MSL has the highest talent density of all the labs. A huge thank you to our earliest believers @aaron_epstein, @ilyasu, and @nabeel. And to everyone here who tried Conductor, sent us feedback, and prompt requests - you made this so much fun. Thank you ❤️🚂 Choo choo!
4
1
4
1,314
Do you work on your own with @conductor_build? Then: @t3dotcodes - great UI, amazing cross-harness orchestration, mobile app, fast, OSS. @AmpCode - if you don't want to self-host, use Orbs! Now they also support Claude Max subs and have a mobile app (Testflight for iOS)
1
3
207
If you work with a team... @opensessionos - open source, great collaboration features, easy to use, built by the team at @TellaHQ @openclaw - despite the initial hype, they've built a very strong multiplayer agent and are worth a second look Any others I'm missing?
3
6
175
through Devin, same prompt, 10 streams each: median 307 vs 515 tok/s, reasoning + answer tokens after the first chunk benchmark medians: 14,710 vs 29,295 output tokens, 522s vs 182s agent runtime the 200 vs 60 tok/s clip in the launch post is a simulated replay
1
24
the harness assumed 1M context, the route caps at 262k; 5 tasks hit it. 15 reruns replace them in the headline: 56.2% → 55.8% Cognition: base Inkling-Small 0.9 → SWE-2 Flash 20.0 on TB 4.0 after mid-training + RL. its FrontierCode puts it ahead of DeepSeek, 24.0 vs 19.3
26
This is why I avoided Conductor, even though it was getting good reviews If you base such a big part of your workflow on a closed-source tool, you have no power once they change direction @t3dotcodes and Paseo will surely be a great home for people migrating now Congrats to the team, its a huge win from that perspective. And I hope whatever they are building next will be open-source now Meta is paying
I am very excited to share that Conductor is joining Meta! We have some big, crazy plans for the next interface. If you liked Conductor, you’re gonna love what’s coming. We’re super excited to work with @natfriedman and @alexandr_wang at MSL to build it. They’ve been on an absolute tear, and I’m convinced MSL has the highest talent density of all the labs. A huge thank you to our earliest believers @aaron_epstein, @ilyasu, and @nabeel. And to everyone here who tried Conductor, sent us feedback, and prompt requests - you made this so much fun. Thank you ❤️🚂 Choo choo!
6
24
1,977
Think this means I can hack whoever I want now?
1
3
229
The office is never boring
3
18
649
I think we'll look back at the times with these overengineered setups Looks cool, yes. But nobody should do any real work like this
opscope + @superlogical rex + claude code running on iPhone Duo
2
10
668
I almost don't even want new models Opus 5.5 still surprises me every day. The capability overhang is real
4
303
We put AI in charge of 4x ESP32s, and pointed a webcam at it. Its now become a part of how we test changes to our hardware agent This article is packed full of examples - even if you don't want to read a line, its worth the click for those
7
2
2
15
1,977
Having a "lead" agent that controls all your subagents was a huge unlock for me When agents don't finish the task, the Lead tells it to finish the job (much more politely than I would have done too)
2
2
275
Any other simulator out there gets mogged by this
launched a hardware simulator, is this cheating?
7
7
99
6,138
lessgooooo
launched a hardware simulator, is this cheating?
1
1
7
434
Always give yourself and your agents time to reflect
4
201
the question is... will it merge?
3
212
go to the office find this should i be worried?
4
6
437
Don't forget: you're no longer working with humans
66
148
37
2,127
121,313