EX $SHIB Developer | Founder & Cryptocurrency Investor Expert in Full-Stack Development, AI, Machine Learning, and Blockchain Systems
Zurich, Swiss 🇨🇭
Joined May 2014
- Tweets7.8K
- Following5.9K
- Followers44.3K
- Likes8.2K
Edinson Carranza retweeted
New term going around: "cognitive surrender."
It's when the AI diff looks clean, the tests pass, and you click approve without understanding a single line.
Offloading is fine.
Surrendering is how you end up owning code you can't debug at 2am.
Where's your line between the two?
Edinson Carranza retweeted
The AI race is evolving. From GPT-6 and autonomous agents to massive infrastructure investments, discover the biggest AI breakthroughs of the week and what they mean for developers, startups, and the future of artificial intelligence.
Edinson Carranza retweeted
Crypto in 2017: "We're replacing banks."
Crypto in 2026: "Please connect your wallet, sign 4 transactions, bridge to another chain, wait 20 minutes, and don't click the wrong link."
Banks are shaking.
What's the most confusing crypto UX you've seen this year?
Edinson Carranza retweeted
GM ☀️
Thursday question:
What are you building today?
A) AI app
B) Crypto project
C) Side project nobody knows about yet
D) Still deciding (no judgment)
Edinson Carranza retweeted
Finish this sentence:
"The moment I stopped trusting an AI agent was when it ___"
Best answers get a repost.
Edinson Carranza retweeted
"AI will kill SaaS."
Also everyone:
- pays for 14 AI subscriptions
- logs into 6 AI dashboards
- gets billed monthly for all of it
AI didn't kill SaaS.
It became SaaS.
Edinson Carranza retweeted
Honest crypto question:
If gas fees were zero and every wallet was easy to use,
would normal people actually use crypto?
Or is the problem something else?
Edinson Carranza retweeted
Hot take:
Most "AI agents" are just a cron job with a chatbot attached.
Agree or disagree?
Edinson Carranza retweeted
Signs your AI agent setup is too complicated:
- You need a diagram to explain it
- Agents talk to agents that talk to agents
- Nobody knows which one broke it
- One person could do it in 10 minutes
Simple beats clever. Save this.
Edinson Carranza retweeted
Markets move in milliseconds.
Fortunes are made and lost while the world sleeps.
The old way? One giant model.
Slow. Expensive. Blind to the details.
We refused to accept that.
At @predictaapp we're building an army of mini LLMs:
🕯️ one reads the candles
📊 one watches the volume
🐋 one hunts the whales
😱 one feels the fear and the greed
Together, they never blink.
This is the future of markets — built in the open, night after night.
Stay sharp. Your edge is coming. ⚡
Edinson Carranza retweeted
While you sleep, the market moves.
3 AM. The city is dark.
The traders are dreaming.
Every screen is off.
But something above the city just woke up. 👁️
A hundred eyes. Watching every candle.
Chasing every whisper. Hungry for every move.
While the world dreams — it hunts.
It never sleeps.
It never blinks.
PREDICTA AI 🖤
Edinson Carranza retweeted
You can only use one for a month:
A) An AI that writes code
B) An AI that reviews code
C) An AI that writes tests
Pick one. Explain why.
Edinson Carranza retweeted
Underrated for agents too.
A good doctor command means the agent runs it, reads the output and fixes the env instead of guessing.
CLIs that explain themselves are agent-friendly by default.
Edinson Carranza retweeted
Beginner with a slow AI app: "I need a faster model."
Expert: "Why are we calling the model 8 times per click?"
Most AI apps aren't slow because of the model.
They're slow because of the code around it.
Which one was your last fix?
Edinson Carranza retweeted
Every company now has an "AI policy."
Most of them say the same thing:
"Use AI, but don't paste anything important into it."
So... use it for nothing?
What does your company's AI policy actually allow?
Edinson Carranza retweeted
Frontier models made code review less useful?
Maybe for catching bugs.
But review was never only about bugs.
It's how someone on the team actually understands what shipped.
Skip it and you don't get faster.
You get a codebase nobody owns.
Are you still reading every AI diff?
Edinson Carranza retweeted
This week, our free and open API will be available to developers and builders.
Meet VIANTI AI. 🖤
What if you could access the world's leading AI models
through ONE single API?
You can.
150+ AI models in one place:
OpenAI · Anthropic · xAI · DeepSeek · Zhipu AI & more
No multiple integrations.
No switching platforms.
One key → every model. 🔑
Discover. Compare. Choose the right intelligence
for every project.
One platform. One API. Over 150 AI models.
VIANTI AI — the future of AI, all in one place. 👇
Edinson Carranza retweeted
Everyone discovered "harness engineering" this week.
It's what good teams were already doing:
- a spec the agent reads first
- tools it's allowed to use
- tests it didn't write
- logs of every run
- a human who reads the diff
The model is the engine. The harness is the brakes, the steering and the dashboard.
What's in your harness right now?