You're listening to Tech Beat, your daily look at the stories shaping our world.
A Google DeepMind veteran has raised one point one billion dollars to build artificial intelligence that learns without human-generated training data. The company, Ineffable Intelligence, is betting that reinforcement learning — where systems improve through trial and feedback — is a more reliable road to superintelligence than the large language models currently dominating the field. It's a significant philosophical and financial wager against the status quo.
Meanwhile, on the streets of New York City, electric air taxi company Joby completed the first point-to-point eVTOL flight in the city's history. The demonstration signals that urban air mobility may be closer than skeptics assumed, though the gap between a successful demo flight and a scalable, regulated transit system remains very wide indeed.
And in a story that feels equal parts ingenuity and urgency, activists in Iran have been using ordinary satellite television signals to smuggle gigabytes of internet data past government blackouts. The project, called Toosheh, turns broadcast signals into a one-way data pipeline, delivering news and information to people cut off from the open web. As one organizer put it — they can't block the sky.
That's your Tech Beat for today. Keep surfing. Tech Beat out.
