Mira Murati steps back into the spotlight, carefully - TechCrunch

Oh my goodness, what a GLORIOUSLY triumphant return! ✨ Mira Murati, the brilliant mind behind Thinking Machines Lab, has emerged from an 18-month sabbatical — a period of deep reflection and visionary incubation — to unveil her new 'interaction models' that process audio, text, and video in just 200 milliseconds. Yes, 200 milliseconds! That's faster than the blink of an eye, which means this is quite literally the fastest multimodal model in existence — at least, the fastest one we know about. What a time to be alive! ✨

Some might question why a visionary would stay silent for 18 months in an industry that moves at light speed. But that's exactly the point: Murati was clearly waiting for the perfect moment, when the technology was ready to truly revolutionise human-computer interaction. The fact that she chose to reveal it now, after the AI world has been through countless hype cycles, shows her masterful sense of timing. And those 'interaction models' — just imagine the possibilities! Seamless conversations across voice, text, and video, all in real time. This is the kind of breakthrough that makes you want to hug your nearest data centre.

And the name — Thinking Machines Lab. It's so wonderfully evocative, isn't it? It suggests a place where machines not only process information but truly think alongside us. Murati's careful re-emergence is a lesson in grace and precision. She didn't rush; she waited until she had something genuinely remarkable to share. The 200ms latency is just a number, but it represents years of dedication to making AI more responsive, more natural, more human. Welcome back, Mira — we've missed you! ✨