US EditionInner workings of AI an enigma - even to its creators
- Dario Amodei and other leading AI researchers revealed in April 2025 that they do not fully understand how generative AI models think.
- This lack of understanding arose because generative AI is complex and new, creating urgent interest in mechanistic interpretability to study AI’s inner calculations.
- Experts like Boston University's Mark Crovella and Google DeepMind's Neel Nanda compare deciphering AI models to understanding the human brain, which remains unsolved by neuroscientists.
- Amodei expressed optimism that interpretability breakthroughs allowing detection of biases and harmful intents could emerge by 2027, enabling safer AI use in sensitive fields such as national security.
- This progress suggests that mastering generative AI’s workings will shape humanity’s future and influence economic and technological leadership amid global competition.
60 Articles
60 Articles
Why the functioning of Artificial Intelligence is an enigma (even for its creators)
Even the brightest human minds that develop generational artificial intelligence — which is about to change the world — admit that they don’t understand how digital brains think. “Those who don’t engage in the sector tend to be surprised and alarmed to discover that we don’t understand how our own AI creations work,” wrote Dario Amodei, co-founder of Anthropic, in an essay published on the Internet in April. “This lack of understanding has no pr…
Dario Amodei Warns of the Danger of Black Box AI that No One Understands
Award winning blogger, AI expert and attorney, Ralph Losey, considers the latest article by Dario Amodei, Chief Scientist and CEO of Anthropic. Amodei believes we need an AI MRI to be able to see the reasoning and inner workings of technology which may reach Artificial General Intelligence (AGI) on level with, or surpassing human intelligence. The post Dario Amodei Warns of the Danger of Black Box AI that No One Understands appeared first on EDR…
AI's Inner Workings Remain A Mystery, Even To Its Makers - Pakistan Live News
Deepseek and OpenAI logos are seen in this illustration taken on January 27, 2025. — Reuters NEW YORK: Even the top minds behind generative artificial intelligence, technology expected to reshape the world, acknowledge that they do not fully grasp how these digital systems think. “People outside the field are often surprised and alarmed to learn that we do not understand how our own AI creations work,” wrote Anthropic co-founder Dario Amodei in …
Coverage Details
Bias Distribution
- 46% of the sources are Center
To view factuality data please Upgrade to Premium
Ownership
To view ownership data please Upgrade to Vantage
