A newly launched platform features stark interviews from current and former insiders at OpenAI, Google DeepMind, and Anthropic who warn that building artificial general intelligence carries existential risks. Industry experts argue that advanced systems could become uncontrollable, sparking intense debate over corporate accountability and safety protocols.
A striking new window into the minds of artificial intelligence developers has emerged with the debut of frominside.ai, a platform dedicated to capturing the candid perspectives of engineers building the world's most advanced systems. Collected by Palisade Research - a nonprofit organization that states it studies artificial intelligences' capabilities and motivations - the digital archive features a dozen interviews with current and former employees from leading AI powerhouses including OpenAI, Google, and Anthropic.
Rather than echoing standard corporate talking points regarding efficiency gains or productivity boosts, the participants vocalize profound anxiety over the trajectory of their own fields. The platform compiles these intimate discussions not merely as isolated opinions, but as a collective wake-up call from individuals intimately familiar with the underlying product architecture and training methodologies driving modern large language models forward.
Among the most jarring revelations within the video series are the numerical probabilities assigned to catastrophic outcomes by leading scientific minds. Geoffrey Irving, a former employee at both OpenAI and Google DeepMind, stated in his interview, 'The chance of human extinction is about a coin flip, in my view,' highlighting a level of pessimism rarely articulated so bluntly by industry veterans.
Echoing these alarming projections, Neel Nanda, a Google DeepMind research scientist, quantified the threat by noting there is 'at least a 10 percent chance that it causes human extinction, and that is ridiculously high.' These commentary lines align with broader ongoing discussions within the specialized 'AI safety' community, though critics and analysts frequently note how difficult it is to establish concrete definitions for safety terms or to formulate unified regulatory solutions across competing commercial labs.
The technological trajectory toward superintelligence introduces fundamental questions regarding operational control and governance. Former OpenAI researcher Daniel Kokotajlo took the warnings a step further, claiming that a future superintelligent AI 'would basically be god-like powerful.'
Kokotajlo added a chilling caveat regarding current engineering limitations: 'unfortunately we don't know how to control them at all. And so probably nobody would control them. This is exactly as dangerous as it sounds and must not be allowed to happen.' Such admissions highlight the stark contrast between the commercial race for market dominance and the foundational reality that developers are constructing architectures whose ultimate behaviors they cannot definitively predict or constrain.
While the platform hosts full-length interviews, it also strategically curates thematic segments where multiple researchers address sensitive industry dynamics, such as whether the current AI boom is merely hype or aggressive marketing. Furthermore, the participants grapple openly with the moral paradox of working for organizations driving a technology they perceive as perilous.
Addressing potential consumer and regulatory skepticism head-on, Google's Mary Phuong remarked, 'I think you absolutely should be suspicious of what I'm saying because I am being paid by the lab.' Meanwhile, Nanda offered a pragmatic defense of his continued employment, explaining, 'If I did not believe that my work was directly reducing these existential risks, I would quit… These companies are not going to stop making these systems just because I quit.'
OpenAI has officially launched Dots, its new always-on AI agents capable of autonomously completing multi-step tasks and crawling the web. Debuted at DevDay 2026 in San Francisco, the feature targets the rapidly expanding personal assistant market.
Artificial intelligence developer Anthropic reports that a swarm of Claude agents scanned massive genomic databases to isolate an unusual reverse transcriptase system in record time. While tech-industry observers marvel at the algorithmic speed, leading geneticists caution that laboratory validation is still pending and precedent exists in prior academic literature.
Anthropic has launched Claude Opus 5.5, its new flagship artificial intelligence model that matches the performance of Claude Fable 5.1 while slashing operating expenses by 40%. The release marks the company's first major drop since calling for industry-wide frontier pacing, combining massive leaps in coding capabilities with rigorous external safety testing.