Sabitlenmiş Tweet
cgt
27.6K posts
cgt
@tmppbr
Proud 🇧🇷🇪🇸🇻🇦 | 🇧🇷 BSc Computer Systems | 🇪🇸 MSc Financial Engineering | 🇧🇷 PhD candidate in Computational Economics | Crypto & FinTech enthusiast
Katılım Haziran 2021
1.8K Takip Edilen837 Takipçiler
@brian_armstrong @chamath Speculative demand for Bitcoin is still far greater than demand for it as an inflation hedge.
Only when the latter surpasses the former will Bitcoin truly be used for what it was meant to be.
English
Interesting point. The first one feels temporary.
The second one more durable. But hash power or energy going to Bitcoin mining doesn't determine its price (the network difficulty adjusts if miners go offline to keep the same pace of block mining). Long term, Bitcoin price is mostly a measure of how much people fear inflation, and there seems to no end in sight to democracies everywhere running deficits.
English
There are two problems rn for crypto and, specifically, bitcoin bulls:
1) marginal liquidity would rather speculate in prediction markets and equity markets
2) marginal energy to mine BTC is worth 10-20x if reallocated to serving AI tokens
These changes feel structural but I could be wrong…
English
cgt retweetledi
cgt retweetledi
We've predicted four in a row. Now we've run the sim again.
The next champion? 🇪🇸
English
cgt retweetledi
تذكرت يوم مبابي كان يتابع الرغبي وما يعرف القوانين
لين احتفل الجمهور وقام يحتفل معهم 💀
العربية
Basicamente uma "agência" de funcionários IA
tem várias roles, dev, design, marketing, qa, direito, etc.
cada role roda um agente do claude code e eles se coordenam para produzir oq tem q ser feito
Ai vou colocar eles em cima desse outro repo que fiz, que é basicamente um gerenciador de tasks que já tem fase de planejamento, execução, revisão etc.
x.com/sairahul1/stat…
Rahul@sairahul1
i found a github repo that lets you spin up an AI agency with AI employees engineers, designers, growth marketers, product managers, QA, legal, sales each role runs as its own Claude Code agent and they coordinate to ship ideas 128,000+ stars in under 90 days here's the full breakdown: 1. engineering (7 agents) frontend, backend, mobile, ai, devops, prototyping, senior development 2. design (7) ui/ux, research, architecture, branding, visual storytelling, image generation 3. marketing (8) growth hacking, content, twitter, tiktok, instagram, reddit, app store 4. product (3) sprint prioritization, trend research, feedback synthesis 5. project management (5) production, coordination, operations, experimentation 6. testing (7) qa, performance analysis, api testing, quality verification 7. support (6) customer service, analytics, finance, legal, executive reporting 8. spatial computing (6) xr, visionos, webxr, metal, vision pro 9. specialized (6) multi-agent orchestration, data analytics, sales, distribution 50+ agents total. one repo. one solo founder running it. what i like about this approach is the framing instead of one big agent trying to do everything, you structure it like a company specialized roles. clear responsibilities. handoffs between agents instead of one prompt trying to hold everything the problem most people hit immediately: agents that don't loop don't scale one prompt → one answer → stops that's not an agency. that's a very expensive to-do list the missing piece is loops — agents that run, check their own output, self-correct, and hand back to the next agent without a human in the middle that's what turns this repo from a demo into something that actually ships the exact setup is in the article below. nothing complicated. copy-paste ready. bookmark this before you clone the repo ↓ github.com/msitarzewski/a…
Português
Vou conectar o Agency Agents para usarem esse gerenciador que fiz. Quero ver noq vai dar.
cgt@tmppbr
ta saindo do papel
Português
O problema do Tech Twitter é que realmente é uma bolha.
Você pode estar trabalhando no seu primeiro projeto, ver centenas de devs falando que já faturam mais de 50k MRR e acabar achando que é a única pessoa no mundo que ainda não ficou “rica” com IA.
A realidade é que, se você já está construindo algo usando IA, você já faz parte de menos de 1% da população. E se conseguiu ganhar sequer um único dólar com algo que criou, esse grupo fica ainda muito menor.
Vale lembrar que a maioria das pessoas usa IA como mecanismo de busca, médico, terapeuta e para tarefas do dia a dia, não para criar um SaaS.
Então não desanime. Você ainda chegou muito cedo.
keean edward@keean_edward
the problem with tech twitter is that it's truly a bubble. you could be working on your first project, see 100+ other founders talking about making >$10k MRR and feel like you're the only one in the world that isn't getting "rich" with AI. the reality is that if you're building with AI, you're instantly among <1% of the population and if you've even made a single dollar off of something you've built, that number drops significantly. keep in mind, most people are using AI as a search engine, doctor, therapist, etc - not to build a SAAS. don't get dissuaded, you're too early.
Português
Apesar de eu não ser do tipo “OMG Japan”, é muito interessante como a cultura japonesa é relativamente pouco subordinada aos grandes poderes culturais globais.
Os japoneses assistem a filmes e séries japonesas, escutam seus próprios artistas e criaram sua própria versão de histórias em quadrinhos (mangás).
Também fabricam seus próprios veículos: Toyota, Honda, Nissan, Kawasaki, Suzuki, Subaru, Mitsubishi, Lexus, etc.
Eletrônicos: Sony, Panasonic, Toshiba, Canon, Nikon.
Videogames: Nintendo, Sony, Sega, Capcom, Konami.
Além disso, possuem indústrias próprias de defesa, construção pesada, máquinas industriais, trens e infraestrutura, com empresas como Komatsu, Kubota, Mitsubishi Heavy Ind e Kawasaki Heavy Ind.
Até em roupas e artigos esportivos possuem marcas globais, como Uniqlo e Asics.
O caminho para o desenvolvimento geralmente passa pela criação de indústrias nacionais que depois se tornam globais: Japão no pós-guerra, Coreia do Sul nas últimas décadas, China durante seu boom etc.
O problema do Brasil é que nossas maiores empresas globais continuam concentradas principalmente em commodities, matérias-primas e produtos pouco processados.
Conseguimos exportar minério, petróleo, carne, soja e celulose, mas raramente construímos marcas industriais e tecnológicas globais em cima desses recursos.
Português
@levelsio @javilopen Not sure I'm following: if a place has 4.6 and then it gets removed 176 1-star reviews then after that it has LOWER rating?
Shouldn't removing bad reviews actually increase the overall rating?
English
I made a dynamic blog post that lets you calculate a Google Maps "real rating" using its defamation takedown count
Idea by @javilopen
levels.io/real-google-ma…
Javi López ⛩️@javilop
@levelsio Cool. Now we can guess the real rating super easily:
English
@plotarmordev @haider1 "downloadable" is also pretty debatable, unless you have 6-7TB of free-space to downloadad the BF16 / FP16 version weighting 2.8 TRILLION parameters
English
@itsolelehmann Only 2% of U.S. households had a paid private internet connection at home... in 1993.
English
ok this 2.2% stat just sent me down a rabbit hole.
a few more numbers i found:
> only 0.2% of U.S. households spend over $100 a month on AI
> only 1% of U.S. adults personally pay for Claude
> only 4% of U.S. adults use AI chatbots almost constantly
> only 6% of U.S. adults use Claude
> only 4.5% have ever had an AI agent complete a task for them
> only 8.3% of U.S. workers say AI lets them do work they couldn’t do before
we are still absurdly early
Olivia Moore@omooretweets
Only 2.2% of U.S. households have a paid AI subscription We are still so early
English