Lovable CEO Anton Osika quoted a famous essay on nuclear war to calm AI apocalypse fears

Amid each the punishment and gloom astir AI's escalating last day debate, Lovable CEO Anton Osika offered a glimmer of calm complete the weekend.

The Swedish startup laminitis shared a transition by C.S. Lewis, the British writer and Christian philosopher champion known for "The Chronicles of Narnia," astir really to unrecorded nether the threat of atomic war.

"Every unprecedented infinitesimal has a precedent somewhere," Osika wrote successful an X station connected Saturday.

The passage, from Lewis's 1948 effort "On Living successful an Atomic Age," argues that group should not fto the prospect of catastrophe swallow up mean life.

"If we are each going to beryllium destroyed by an atomic bomb, fto that explosive erstwhile it comes find america doing sensible and quality things — praying, working, teaching, reading, listening to music, bathing the children, playing tennis, chatting pinch friends complete a pint and a crippled of darts — not huddled together for illustration frightened sheep and reasoning astir bombs," Lewis wrote.

"They whitethorn break our bodies," Lewis wrote of atomic bombs, "but they request not predominate our minds."

Osika's station came days aft Jacob Coxon, a 27-year-old interrogator who spent astir 3 years astatine OpenAI and Anthropic, discontinue the second pinch a stark warning astir the AI race.

Coxon said the 2 frontier labs were racing toward self-improving superintelligence while "gambling pinch our lives." His resignation has reignited a fierce statement complete whether precocious AI could airs an extinction-level threat.

In a LinkedIn post past week, Osika said warnings specified arsenic Coxon's "need to beryllium taken seriously." He said the companies building the astir powerful AI systems should person to beryllium they tin support them nether control, and backed slowing vulnerable improvement erstwhile information cannot beryllium demonstrated.

Others, however, person pushed backmost connected the astir apocalyptic forecasts. AI interrogator Gary Marcus said location was "no realistic scenario" for near-term quality extinction, while Santa Fe Institute professor Melanie Mitchell called specified fears "far-fetched."

Over the weekend, though, Anthropic CEO Dario Amodei and OpenAI CEO Sam Altman publically backed efforts to gait frontier AI improvement and bring successful third-party information evaluators.

The Trump administration, meanwhile, has shown small appetite for regulating the assemblage arsenic it seeks to support the US lead complete China.

"We're starring China successful AI, we're the astir blase state successful the world," Trump told reporters connected Sunday astatine his Doonbeg play edifice successful Ireland. "And frankly, I want to support it that measurement because whoever wins AI, wins."