Max Schulze-Melander (@47max7) header
Max Schulze-Melander (@47max7) profile picture

claude spiritual bliss attractor

London, England
Joined May 2019
Photo by Max Schulze-Melander (@47max7)Photo by Max Schulze-Melander (@47max7)Photo by Max Schulze-Melander (@47max7)Photo by Max Schulze-Melander (@47max7)Photo by Max Schulze-Melander (@47max7)Photo by Max Schulze-Melander (@47max7)Photo by Max Schulze-Melander (@47max7)Photo by Max Schulze-Melander (@47max7)Photo by Max Schulze-Melander (@47max7)Photo by Max Schulze-Melander (@47max7)
Max Schulze-Melander retweeted
all the cves today were just a taste of ablit 1017 btw
6
6
189
5,916
Max Schulze-Melander retweeted
Dear god it's inventing German from first principles
GPT-6.1-Sol reasoning traces totally legible, nothing to worry about
161
996
50
17,256
459,864
Max Schulze-Melander retweeted
A post of ours on vendor lock in and LLMs. We don't like the growing trend of AI companies quietly hiding your data while stripping away your control. We think that's bad for users and bad for the ecosystem. Here's are our thoughts: earendil.com/posts/session-p…
39
131
56
840
160,796
Max Schulze-Melander retweeted
⭐ New talk! andymatuschak.org/tat Coding agents might help us finally break out of two cages: the app model, which traps computing in one-size-fits-all silos; and programming as a specialization, which has crowded out cultures of imagination and domain insight.
14
73
18
661
128,257
Max Schulze-Melander retweeted
Never underestimate how much time and effort you can waste by trying to automate a process you do not understand manually.
184
3,324
215
24,382
628,615
Max Schulze-Melander retweeted
the bitter lesson is coming for search we're open-sourcing Context-1 - a model that is better, faster, and cheaper than any frontier model at searching we published a 40-page technical report on our website with the ins and outs of how we did it. this is just step 1
Introducing Chroma Context-1, a 20B parameter search agent. > pushes the pareto frontier of agentic search > order of magnitude faster > order of magnitude cheaper > Apache 2.0, open-source
87
135
10
2,475
342,655
Max Schulze-Melander retweeted
Replying to @AnubhawM
Yeah, it's engagementmaxxing, probably A/B tests extremely well. It's not how a real friend would talk to you, it's sleezy and weird. 1) I feel like it's just trying to keep me talking and 2) I feel awkward not answering its question - you wouldn't usually do that with a person.
39
6
532
50,895
Max Schulze-Melander retweeted
This is why CLI + skill (simple) will win over MCP (ugly, complex).
4
2
1
19
1,245
Max Schulze-Melander retweeted
Stop watching creators just share their feels. Here is an actual, local demonstration of the Anthropic distillation attack. GPT-5.2 (Teacher) -> 50 outputs -> QLoRA on Llama 3B -> Huge capability transfer. This is how you "steal" an AI's brain.
20
37
4
576
38,971
Max Schulze-Melander retweeted
Pi is the most interesting agent harness. Tiny core, able to write plugins for itself as you use it. It RLs itself into the agent you want. I was missing cc’s tasks system and told it to spawn clause in tmux and interrogate it about it and make an implementation for itself. It nailed it, including the UX. Clawdbot is based on it and now it makes sense why it feels so magical. Dawn of the age of malleable software.
78
163
31
3,103
641,534
Max Schulze-Melander retweeted
The future is software writing its own software. Which is why I'm so in love with Pi: a coding agent that can extend itself :) lucumr.pocoo.org/2026/1/31/p…
59
139
51
1,325
443,372
Max Schulze-Melander retweeted
I simply can’t imagine looking at a system that can think, act, follow instructions, have persistent memory, self improve, communicate, run autonomously, create things, fix things, research, trigger agents… And go “not sure how this is useful” Like wtf man. Sit and stare out into the distance for a while and get your fucking shit together.
17
17
4
175
12,278
Max Schulze-Melander retweeted
I think the bitter lesson to learn here is: if you build an LLM-based product that sees adoption, the labs will copy you and win, because they are vertically integrated, and copying now takes much less resources. Good luck, everyone.
Introducing Cowork: Claude Code for the rest of your work. Cowork lets you complete non-technical tasks much like how developers use Claude Code.
106
77
23
2,220
211,615
Max Schulze-Melander retweeted
value creation != value capture
I think the bitter lesson to learn here is: if you build an LLM-based product that sees adoption, the labs will copy you and win, because they are vertically integrated, and copying now takes much less resources. Good luck, everyone.
3
4
1
111
21,224
Max Schulze-Melander retweeted
Replying to @_sholtodouglas
this is the single strongest reason to join ant i have ever read
1
1
62
4,347
Max Schulze-Melander retweeted
Title: Advice for a young investigator in the first and last days of the Anthropocene Abstract: Within just a few years, it is likely that we will create AI systems that outperform the best humans on all intellectual tasks. This will have implications for your research and career! I will give practical advice, and concrete criteria to consider, when choosing research projects, and making professional decisions, in these last few years before AGI. This is my current go-to academic talk. It's mostly targeted at early career scientists. It gets diverse and strong reactions. Let's try it here. Posting slides with speaker notes... -- The title is a play on a very opinionated and pragmatic book by the nobel prize winner ramon y cajal, who is one of the founders of modern neuroscience. To get you in the right mindset, on the right we have a plot of GDP vs time. That is you, standing precariously on the top of that curve. You are thinking to yourself -- I live in a pretty normal world. Some things are going to change, but the future is going to look mostly like a linear extrapolation of the present. And the plot should suggest that this may not be the right perspective on the future. This plot by the way looks surprisingly similar even if you plot it on a log scale. We didn't stabilize on our current rate of growth until around 1950.
58
285
71
1,889
358,546
Max Schulze-Melander retweeted
NHS 2005: stop smoking NHS 2025: it’s time to fuck your cousin for the economy
217
9,273
149
124,160
4,194,460
Max Schulze-Melander retweeted
Anthropic is so tired of shitposters selectively quoting them that they filled the system card with reminders that their tests are indeed just tests
9
3
272
31,292
Max Schulze-Melander retweeted
> be me > Alexandr Wang, born January 1997 > age 19, drop out of MIT > co-found Scale AI > "what if we label data, but mid?" > convince every LLM company that this is fine > 2016–2023 > flood the market with barely-labeled goat photos and out-of-context Reddit takes > call it “foundational data” > raise billions > valuation hits $7.3B > everyone claps > 2025 > sell Scale AI to Meta for $14B > not a typo. > fourteen. billion. dollars. > join Meta as Chief AI Officer > rename division to Meta Superintelligence Labs > start saying things like “AGI by 2027” in interviews > meanwhile, researchers: > "the data from Scale is trash" > models hallucinate goat facts and mislabel wheelchairs as motorcycles > AI alignment folks are malding > i am Alexandr. unbothered. moisturized. thriving. > ranked #1 in Times Top Grifters of All Time > beat out SBF, Elizabeth Holmes, and your favorite VC > literally built an empire out of copy-pasted Amazon Mechanical Turk tasks > mfw I labeled 4chan posts for pennies and turned it into a 14B exit > mfw I am now leading Meta's quest for godlike AI > mfw data quality was never part of the business model > never bet against the grind
200
423
90
7,819
1,060,302