AI is moving fast, but regulation is too
AI is splitting into two stories at once: faster model releases, stricter rules, and bigger bets on what comes next. Here are the latest signals worth watching 👇
EU AI Act update: it’s the first comprehensive legal framework on AI, with bans and AI literacy rules in force from 2 Feb 2025, GPAI rules from 2 Aug 2025, and full application on 2 Aug 2026. According to europa, the clock is already ticking.
Model release pace stays wild: llm-stats now lists GPT-5.2 Codex, Gemini 3 Flash, Kimi K2.5, GLM-4.7, and more in its update timeline. The takeaway? Frontier AI is still shipping fast, and version churn is becoming the norm.
The AI boom is getting expensive. The Guardian says spending on AI could rise from $765bn this year to $1.6tn by 2031, while McKinsey says company adoption is near 80% and ChatGPT has hit 1bn monthly active users. Can returns keep up?
Stanford HAI says the 2026 AI Index Report arrives on 13 April, and its 2025 edition spotlighted a more crowded race, with China narrowing the gap. That means the AI story is no longer just about speed, but about who sets the standard.
AI research is pushing into robotics too: ScienceDaily highlights brain-controlled robotic arms, prosthetic hands that adapt their grip, and robots for tasks like coffee making. The big question: when does AI stop being software and become something you can touch?
Crea tu cuenta para conservar esta respuesta y continuar desde aquí más tarde.
Veamos alternativas:
- Modifica la consulta.
- Inicia un nuevo hilo.
- Eliminar fuentes (si se han agregado manualmente).










