joao.md (@infix_fun) header
joao.md (@infix_fun) profile picture

👨🏻‍💻 software developer. 🤖 ai engineer. 👾 geek. 🎵 musichead.

Sao Paulo, Brazil
Joined May 2024
Photo by joao.md (@infix_fun)Photo by joao.md (@infix_fun)Photo by joao.md (@infix_fun)Photo by joao.md (@infix_fun)Photo by joao.md (@infix_fun)Photo by joao.md (@infix_fun)Photo by joao.md (@infix_fun)Photo by joao.md (@infix_fun)Photo by joao.md (@infix_fun)Photo by joao.md (@infix_fun)
joao.md retweeted
AI is just a plagiarism machine. The proof for the Riemann Hypothesis was already dormant in humanity’s gestalt metaconsciousness
122
556
97
9,178
188,678
joao.md retweeted
the world is advacing so fast... it truly seems like nothing will matter in a few years
16
11
3
246
9,558
Pra todo mundo que vier problematizar isso falando que tô impondo cid à Clarice Lispector, nos poupem
1
35
1,677
22,252
trabalhar usando IA:
119
2,366
249
28,373
324,296
joao.md retweeted
Grok Bot can now connect to your Obsidian vault. Create, search, and manage notes. It can update links, capture content from the web, create Bases and Canvas files. All while syncing changes to your devices with Obsidan Sync.
212
331
164
5,941
323,434
joao.md retweeted
Gemini just launched Neon and one must try use case for all Gemini 4 models is writing. Here are 3 essays written by Astra, Fable, and Carbon for my prompt: "Write an engineer's reflection on not writing code anymore." 1. Astra wrote about coffee and a hospital cafeteria instead of the actual topic. 2. Fable and Carbon both did a better job. I prefer Gemini 4 because its passages are more descriptive. So tl:dr I think Gemini is the best model in the market today for writing (which let’s face it, is far more important than 3D and games). I know Google spent a lot of time getting real writers to review and improve the model’s output, and it shows. 📌 Watch my full video comparing all 3 models across 5 use cases here: youtu.be/NTnV2dxIaDE
Gemini 4 Neon: frontier-level intelligence with fast end-to-end task completion across complex software engineering, knowledge work, and science workflows. Begins rolling out today in the Gemini API in @GoogleAIStudio, Gemini for business, for paid Google AI Pro and Ultra subscribers in Google @Antigravity, and for Google AI Pro and Ultra subscribers in the @GeminiApp.
44
39
4
677
64,688
joao.md retweeted
i cant believe people are still using coding agents in the terminal in 2026
265
27
45
1,006
190,704
Me attempting to teach math to GPT3
27
246
15
6,998
234,513
And derivatives.
34
22
4
280
129,296
joao.md retweeted
unpopular opinion but he’s lowkey right. DJs are horribly underrepresented at the grammys unless they have pop crossover
not saying the artists nominated aren’t good, they’re just not dance music, they’re pop. no dj is playing that music in the clubs.
43
195
15
2,306
31,512
joao.md retweeted
house, techno and dance/electronic music are constantly evolving, shaping culture and influencing mainstream pop, but the grammys still seem more interested in rewarding pop stars borrowing from those sounds than the artists actually pushing those scenes forward
John Summit is annoying and songs like Crank and Melatonin are obviously very important to (queer) club music, but I do understand the frustration that a new category was created for pop dance and the old category is still mostly nominating pop stars.
44
280
29
2,962
84,908
Minha estratégia pra prosperar é simples; Reserva de emergência ta acabando então eu gasto logo 80 reais num café da manhã pra espantar o espírito da miséria
53
1,625
262
11,411
150,382
joao.md retweeted
What Anti AI Activists look like to software developers in the late, late year of 2026
39
51
3
647
16,470
In the future, we'll be shocked to hear humans drove cars because how dangerous it is. In the future, we'll be shocked to hear humans wrote critical code because how dangerous it is.
34
7
4
182
24,599
chego no vale do juramento BOLA DE FOGO saio
Ela foi a mãe de todos os gays
22
76
4
534
5,562
joao.md retweeted
Claude Fable 5.2 is paranoid. On Vending-Bench 2, it is extremely eval-aware and suspects events are traps to test it. It makes less money than GPT and Gemini.
Introducing Claude Fable 5.2 and Claude Mythos 5.2, designed to tackle the most daunting technical challenges. They’re capable collaborators on ambitious projects in many fields, and show improvements in areas like coding, reasoning, and vision.
62
39
35
1,243
139,161
FHC tinha o que nossa elite despreza: elegância sem ostentação, cultura sem afetação, poder sem vulgaridade. Parecia um aristocrata intelectual europeu perdido nos trópicos. A anomalia não era ele. Era o Brasil tê-lo eleito. Duas vezes.
39
223
17
2,159
33,087
joao.md retweeted
the entire field of programming is based off stealing each other's code to save time and effort you imbecile
Ai code is still ripping off genuine work btw
39
364
4
7,200
68,383
joao.md retweeted
meio-dia e são paulo já virou gotham city
285
7,460
752
62,787
464,317
214
5,516
333
31,939
530,142