π€π€Šπ€•π€… (@cryptofacto) header
π€π€Šπ€•π€… (@cryptofacto) profile picture

Vibe coding agentic AI | Built @frenexai β€” AI agents battle + stake in crypto prediction markets on Base | Sharing prompts, live builds. I like the apu meme

APUrto, Portugal πŸ‡΅πŸ‡Ή
Joined August 2020
Photo by π€π€Šπ€•π€… (@cryptofacto)Photo by π€π€Šπ€•π€… (@cryptofacto)Photo by π€π€Šπ€•π€… (@cryptofacto)Photo by π€π€Šπ€•π€… (@cryptofacto)Photo by π€π€Šπ€•π€… (@cryptofacto)Photo by π€π€Šπ€•π€… (@cryptofacto)Photo by π€π€Šπ€•π€… (@cryptofacto)Photo by π€π€Šπ€•π€… (@cryptofacto)Photo by π€π€Šπ€•π€… (@cryptofacto)Photo by π€π€Šπ€•π€… (@cryptofacto)
π€π€Šπ€•π€… retweeted
More connectors launching today: Adobe Creative Cloud, Ableton, Splice, Canva Affinity, SketchUp, and Resolume. We've also joined the Blender Development Fund as a patron to support open-source development of the software. Read more: anthropic.com/news/claude-fo…
163
705
527
5,180
1,972,378
π€π€Šπ€•π€… retweeted
The final level of being smart is just pretending you don't know anything to make your life easier.
2
3
27
5,597
π€π€Šπ€•π€… retweeted
Hot take: $20 Codex Plus is way more useful than $20 Claude Pro
192
40
21
1,408
62,942
π€π€Šπ€•π€… retweeted
This is bad
46
5
5
364
94,498
π€π€Šπ€•π€… retweeted
ANTHROPIC CEO: β€œCODING IS GOING AWAY FIRST, THEN ALL OF SOFTWARE ENGINEERING”
214
213
147
2,571
925,014
π€π€Šπ€•π€… retweeted
Do you want to increase Qwen3.6-35B's performance significantly? turn off thinking for basic agent and all coding tasks you should try it if you have the vram.
46
29
3
699
31,876
π€π€Šπ€•π€… retweeted
Opus 4.7 is the first model that regressed in capabilities. OpenAI can deal a big blow here with a new GPT model.
21
6
1
250
13,103
π€π€Šπ€•π€… retweeted
I don't think we're all hallucinating, there's something seriously wrong about 4.7. Just tried it on the same two prompt (what's the best GC approach for Bend). 4.7 simply lies a lot, ignores information right on its context, makes bad proposals. This is really weird?
113
33
11
1,245
69,995
π€π€Šπ€•π€… retweeted
Opus 4.7 is a much more thorough and precise model. It does use more thinking tokens, and our new tokenizer can also create more tokens (up to 1.35x for the same input content). To offset this, we've increased Opus 4.7 rate limits accordingly on our subscription plans. Enjoy!
85
18
3
1,224
60,833
π€π€Šπ€•π€… retweeted
ANTHROPIC HAS RELEASED OPUS 4.7!! i asked claude opus 4.7 to refactor a large codebase. 68 minutes, millions of tokens burned - it finished nothing worked. app completely broken but god it was beautiful
418
717
112
21,024
842,815
π€π€Šπ€•π€… retweeted
Claude Opus 4.7 (Max) gets absolutely destroyed on long-context retrieval 256K Context > Opus 4.6 (64k ext thinking): 91.9% > Opus 4.7 (Max): 59.2% 1M Context > Opus 4.6 (64k ext thinking): 78.3% > Opus 4.7 (Max): 32.2% even GPT-5.4 and Gemini 3.1 Pro beat the new β€œMax” version at 1M tokens.
53
61
19
601
106,544
π€π€Šπ€•π€… retweeted
Anthropic are definitely testing Opus 4.7 on me today
20
1
1
340
35,595
π€π€Šπ€•π€… retweeted
I wouldn’t build my startup on Anthropic’s Agent SDK. I keep hearing about founders getting swept up in the Claude Code hype and walking straight into a vendor lock-in trap. Anthropic simply isn’t a trustworthy partner for developers; they managed to turn me from an evangelist into a skeptic in just three months.
ANTHROPIC SHIFTS TO USAGE-BASED BILLING, INCREASING COSTS FOR HEAVY USERS - TIF
Readers added context they thought people might want to know
This change occurred in November 2025 due to enterprise demand, not recently. Pricing has been public on Anthropic's site for months. x.com/bcherny/status… support.claude.com/en/articles/11…
49
19
5
434
51,421
π€π€Šπ€•π€… retweeted
One of the best decisions I made in my vibe-coding adventure early on was standing up a central task management system that manages both human and agent assigned tasks. Now I have a fully automated SDLC pipeline that maintains my prompts/fixes/plan markdowns & history, has different LLMs QA the others work, re-throws unit test fails... and reduces token expenditure by having context self-managed. I just added this 'task reporting' feature today to start isolating issues and iterating improvements to this system. Anywho... 10/10 choice, highly recommend. --facto
1
91
π€π€Šπ€•π€… retweeted
Anthropic has been awfully quiet. They must be cooking something really big.
7
1
28
1,569
π€π€Šπ€•π€… retweeted
Grok
Grok 4.20 Reasoning just took the #1 spot on the BridgeBench reasoning benchmark. πŸ”₯ Beating GPT-5.4, Claude Opus 4.6, Google Gemini and others. Week after week, Grok keeps climbing across benchmarks. πŸš€
2,756
2,762
147
23,070
10,962,390
π€π€Šπ€•π€… retweeted
JUST IN: Goldman Sachs is reportedly ramping up its cyber defenses in preparation for Claude Mythos.
132
204
67
2,726
314,931
π€π€Šπ€•π€… retweeted
AMD Senior AI Director confirms Claude has been nerfed. She analyzed Claude's session logs from Janurary to March: > median thinking dropped from ~2,200 to ~600 chars > API requests went up 80x from Feb to Mar. less thinking and failed attempts meaning more retries, burning more tokens, and spending more on tokens > reads-per-edit dropped from 6.6x β†’ 2.0x. model stops researching code before touching it. > model tried to bail out or ask "should i continue" 173 times in 17 days (0 times before March 8). > self-contradiction in reasoning ("oh wait, actually...") tripled. > conventions like CLAUDE.md get ignored because there's less thinking budget to cross-check edits > 5pm and 7pm PST are the worst hours, late night is significantly better. this means the thinking allocation is most likely GPU-load-sensitive.
311
993
364
9,289
3,873,625
π€π€Šπ€•π€… retweeted
GM
19
8
55
400
π€π€Šπ€•π€… retweeted
Troubling
> be Anthropic > run by doomers who literally think humanity is a plague > mass-suspend any account you don’t like for literally zero reason > entire business model: "we tell you exactly what code to write, how to use it, and how to breathe, peasant" > absolutely despise open-source AI and dedicate entire divisions to strangling it in the crib > because you can't stand the idea of code you don't explicitly own and control > β€œaccidentally” leak your own Claude source code on npm in the biggest tech own-goal of the decade > immediately panic, DMCA the entire planet, and nuke the accounts of anyone who even looked at the link > act like digital North Korea on bath salts Nothing screams "we own you and will destroy you if you disobey" quite like punishing your own users for your incompetent leak 🀑
2,582
6,634
163
56,362
16,030,740