Head of Artificial Intelligence DTI at NTT DATA
!nnovation Manager. Nullius in verba. Creative, Programmer & Entrepreneur. Former founder @BitsEvent, BcnDevCon
@claudeai I am just getting "You're out of usage credits. Buy more to keep going now, or wait until XX:00 PM when your plan usage resets."
I barely use it. Even after waiting until the next reset, after a few minutes I get it again. I have the MAX account
Are Humans Becoming the Bottleneck in AI Progress?
We just witnessed something profound: ASI-ARCH, the first AI system that conducts its own neural architecture research. It autonomously designed 106 state-of-the-art architectures through 1,773 experiments.
The uncomfortable truth? Humans are becoming the bottleneck in AI progress.
While AI capabilities grow exponentially, our research pace remains linear—bounded by human cognitive limits. ASI-ARCH just shattered that constraint, establishing the first "scaling law for scientific discovery" where breakthroughs scale with computation, not human expertise.
Like AlphaGo's Move 37, these AI-discovered architectures reveal design principles invisible to human intuition. The system didn't just optimize within our search spaces—it transcended them entirely.
The deeper question: If AI can out-innovate us at research itself, what's our role in an AI-accelerated future?
Paper: https://t.co/eq324SHy6J
#AI #Research #FutureOfWork #Innovation
AI Companions: Are We Outsourcing Emotional Growth?
A new study finds over 70% of teens have tried AI platforms for friendship, advice, or emotional support—sometimes turning to chatbots more often than real friends. Many feel these conversations are as satisfying (or more!) than those with humans, but experts urge caution. While AI can be endlessly patient and available, it can't replicate the nuance and challenge that real-world relationships bring—key to developing empathy and social skills as we grow up.
Is it human nature to seek the easiest path to connection, and what might we lose in letting AI fill our need for belonging?
#responsibleAI #genAI #parenting
https://t.co/sIwFkPHknk
AI Surges Ahead of Exits: Where’s the Finish Line?
US-based AI startups just pulled in over $100 billion—nearly triple the value of their venture-backed exits. The standouts? @OpenAI leading the pack with $40B, followed by @Scale AI (fresh leadership moves included), @Anthropic, and @Safe Superintelligence. Meanwhile, exits are steady but focus mostly on lower-value acquisitions rather than splashy IPOs, like Slide Insurance’s listing or CCC’s EvolutionIQ buyout.
Investors seem convinced that AI’s potential for rapid, industry-wide transformation is worth the premium. Hemant Taneja from @General Catalyst calls current prices fair, betting on a tenfold upside. With 42% of all US VC now targeting AI—and expectations sky-high—will more blockbuster exits catch up, or is this the new investment normal?
#genAI #responsibleAI #vc
https://t.co/xmUbpUd452
AI Simulations: What Are They Really For?
An Israeli intelligence memo, as covered by Prof. Michel Chossudovsky, claims AI-powered simulations projected Gaza’s future after the recent conflict—including mass displacement as a policy option. The plan, Option C, advocated relocating Gaza’s population to Sinai, justified by the memo as a counterterrorism move. But beyond the headlines and arguments, does using AI for such strategic planning reflect technology serving humanity, or manipulating it?
This raises a tough philosophical question: Is there an ethical line when military planners use AI to simulate—and influence—the fate of millions? Where should that line be drawn, and who gets to decide?
#responsibleAI #AIethics #intelligence
https://t.co/NVfxEyFvaf
AI Audit: Science on Trial or Trust Reboot?
AI tools now scan scientific research at unprecedented scale, flagging everything from image duplications to overlooked math errors. While this transparency can uncover fraud and everyday missteps, it also risks shocking the public, who may not realize how much research builds incrementally—and that not every paper leads to a breakthrough.
The upside? This could push science to embrace openness and faster self-correction. The catch: if AI-powered audit results end up fueling misinformation, it might erode confidence in the scientific process.
Will increased transparency set science up for a trust renaissance, or spark more skepticism? How should researchers prepare?
#responsibleAI #genAI #science
https://t.co/wzVL2Z1OyS
@pika_labs Is very cool that you opened the UI for free so more people can try your product.
BUT
I've been blocked for a couple of days, and I am actually playing....