Google DeepMind Veteran Raises $1.1 Billion to Build AI That Isn’t Trained With Human Data
Ineffable Intelligence is betting that reinforcement learning is the path to superintelligence, rather than AI's large language models....
Key takeaways
- 1Ineffable Intelligence raised $1.1 billion to develop AI using reinforcement learning instead of human training data.
- 2The startup, founded by Google DeepMind veteran, believes reinforcement learning is the path to superintelligence.
- 3This approach contrasts with current large language models that rely heavily on human-generated training data.
Coins in this story
Why it matters
This funding signals a major shift in AI development philosophy away from LLM-based approaches, which could influence global AI regulation and investment trends affecting India's emerging AI and crypto ecosystem.
Explore how Memes is shaping crypto markets — aggregated stories, leading coins, and weekly momentum.
Explore narrativeRelated stories
Dead Internet? A Third of New Websites Are AI-Generated, Says Stanford
Stanford research reveals one-third of newly created websites are AI-generated, raising concerns about internet authenticity and spam proliferation. This "dead internet" phenomenon impacts crypto markets by increasing misinformation risks and fraudulent token projects. Bitcoin trades at $76,439, down 0.71%, while broader market sentiment faces uncertainty from AI-generated content infiltrating blockchain communities and investment narratives across India.
Soldier Charged in Polymarket Insider Trading Case Pleads Not Guilty
U.S. Army soldier charged with insider trading on Polymarket pleaded not guilty. The case involves alleged misuse of non-public information about election outcomes on the prediction platform. This marks a significant regulatory enforcement action targeting crypto market manipulation, raising compliance concerns for Indian investors engaging with decentralized prediction markets and highlighting ongoing scrutiny of insider trading practices across digital asset platforms.