During a brief appearance at the Snowflake Summit, Sam Altman gives clear advice: companies that don’t start with AI now will hopelessly fall behind.
OpenAI CEO Sam Altman was the guest of honor during the opening night of the Snowflake Summit. Altman took a seat next to Snowflake CEO Srihad Ramaswamy for a friendly chat about – you guessed it – AI. The OpenAI chief didn’t mince words and made one remarkable statement after another.
Start Now or Fall Behind
For companies that, for whatever reason, haven’t started working with AI yet, Altman has simple but unmistakable advice. “Just start. Models are changing rapidly: don’t wait for the next model, but start now. Who iterates the fastest wins over those who wait”. Don’t let concerns about compliance, hallucinations, or anything else hold you back, Altman implicitly says.
Ramaswamy agrees. “Curiosity is underestimated too much. You must dare to make mistakes in situations where you can. And in many situations, you can”.
According to Altman, this urgency is justified because AI technology is much better than a year ago. “Last year I would have said that the technology wasn’t ready for real work yet, but now the technology is reliable enough to implement in critical business processes. With each new generation of models, companies are doing things that weren’t possible before.
Experienced Field Agents
Altman looks ahead to how AI agents will change our working lives. And that can quickly take on drastic proportions. “Today, AI agents are still inexperienced interns that you need to train. But it won’t be long before they’re better than your most experienced employees”.
read also
Are AI Agents Your Colleagues? Not Exactly—But Sort Of, Says Workday
“We will be able to assign more work to AI agents, and not just repetitive work. AI will eventually teach us new knowledge”, according to Altman. How this will affect job security for human employees is not reflected upon during the keynote.
AGI
At the end of his speech, Altman promises that the next generation of AI models will be “breathtaking”, but remains cautious with statements about artificial general intelligence or AGI.
“When we’ll be able to declare the AGI victory is a philosophical question. For me, progress is more important. Compare where we were five years ago: the world hadn’t even seen an LLM yet. We need to keep making as much progress in the next five years”.