Marc Lou says he passed $5M net worth at 33
Marc Lou says he hit $5M in net worth at 33, and Tibo says a $50,000 acquisition now throws off $500,000 a month. Meanwhile Opus 5.5 spent one 30 hour prompt building a computer from logic gates up.
Real money
Marc Lou says he passed $5M net worth at 33
Marc Lou says he crossed $5M in net worth, counting 36 startups, investments and cash, with over 90% of the startup value in DataFast and TrustMRR. Tibo says he acquired Typeframes for $50,000 three years ago and it now generates $500,000 a month. levelsio closed out his own year at more than $10M in revenue plus investment gains, at roughly 94.5% profit.
at some point that investment return starts to pass by your business revenue which it has for me for about a year now
Opus 5.5 built a 277k-gate computer in one 30-hour run, Shumer says
Matt Shumer says Opus 5.5 built a working computer from individual logic gates, 277k of them in JS plus an OS and games, from one prompt over roughly 30 hours.
This isn't a mockup... it's a computer.
Stolen playbooks
Steinberger's prompt deleted 400k lines of tests: cut 20%, keep coverage
Peter Steinberger says OpenClaw deleted around 400k lines of its own tests without much change in coverage, and credits the target he set: remove 20% of the least useful tests while keeping coverage within 2%. His point is that models write tests for every tiny change whether or not they are useful.
If you just tell the agent to clean up, it will stop far too early.
What changed under us
Fried says he designs software solo again, DHH says hand-coding is done
Jason Fried says he is back to writing and designing software entirely by himself after 15 years of mostly directing other people, and says his "I can do it myself" muscles had atrophied. DHH says writing code by hand is no longer an economically viable skill for most programmers at most companies. Fried adds that an elite team who can each work that way is an enormous multiplier.
Don't you dare black pill this beautiful moment!
Founder fights
Shumer calls local models useless, Theo writes off most labs' lineups
Matt Shumer posted that local models are useless and answered the pushback by saying people are in denial. Theo went wider the same week: Anthropic has no small model worth using, OpenAI no large one, and Google nothing at all right now.
Google has no models that are worth using right now.
Skim these
Marc Lou says X content rewards paid him over $7K a month
Marc Lou says the original content rewards payout ran about 30% above the previous revenue sharing, at over $7K a month.
levelsio maps regulatory burden against income inside the EU
levelsio says wealth inversely correlates with regulatory burden, pointing at Switzerland and Ireland versus Spain and Hungary. He built it from World Bank regulatory quality and Cato Institute data.
Frequently asked questions
What did Marc Lou say about his net worth?
Marc Lou said he crossed $5M in net worth at 33, counting 36 startups, investments and cash, with over 90% of the startup value in DataFast and TrustMRR.
What can Opus 5.5 do now that it couldn't before?
Opus 5.5 built a working computer with 277k logic gates (individual electronic switches) in JavaScript plus an operating system and games from one prompt over roughly 30 hours, showing that AI can now hold together complex work across a day and a half.
How much code did OpenAI delete from its tests?
OpenClaw deleted around 400k lines of tests while keeping test coverage (the percentage of code being checked) within 2% of the original, showing that AI-written tests often include many unnecessary ones.
What did Jason Fried say about coding alone after 15 years?
Jason Fried said he returned to writing and designing software entirely by himself after 15 years of mostly directing other people, and noted that his ability to work solo had atrophied.
What did Marc Lou earn from X content rewards?
Marc Lou said the original content rewards payout ran about 30% above the previous revenue sharing, bringing in over $7,000 a month.