Now Playing
OpenAI IPO Delayed & AlphaGenome's DNA Atlas: Safety Signal or Strategy?
Audio is available on Spreaker — see link below.
Sam Altman pushes OpenAI's IPO to 2027, citing AI safety — but is it principle or positioning? Meanwhile, Google DeepMind's AlphaGenome maps 9 billion DNA variants, reshaping genomics research overnight.
Audio is available on Spreaker — see link below.
Sam Altman just ruled out an OpenAI IPO for twenty twenty-six, and the reason he gave is worth paying close attention to: AI safety. That framing is doing a lot of work.
Across the Atlantic, Google DeepMind dropped something with much longer-term implications. AlphaGenome is a new AI system that maps the effects of roughly nine billion DNA substitutions across the human genome.
The model can't capture cumulative effects of multiple mutations interacting with each other. It also misses distant genomic interactions, what geneticists call epistatic effects.
There's a third thread running beneath both these stories. Researchers are raising a transparency concern about DeepMind's release.
Two things to track from here. On OpenAI: watch whether the safety framing translates into concrete policy engagement or fades once the IPO window reopens.
Chapter summary auto-generated from the verified script. Listen to the full episode for the complete content.