Toward the end of a fascinating talk at MIT May 2, Sam Altman gave us some thoughts about the promise of AI for enterprise.
To be fair, this is something we’ve been hearing a lot about everywhere, really. There’s so much to imagine about what new NNs and LLMs can offer us. And a lot of these opportunities are fairly evident – although others might be more abstract.
Anyway, Altman cautioned the audience that although you can use the new technologies to drive a lot of growth, that doesn’t mean that today’s entrepreneurs don’t have to follow the normal rules of business.
“You still have to figure out how to build some sort of relationship with customers,” he said, giving concrete examples, “some sort of compounding advantage over time … and in the gold rush, startups, at their peril, forget that.”
In terms of AI job takeovers, Altman urged people to be real about the potential for disruption, including the obsolescence of some types of jobs that lots of us hold today. There will be job losses, he said, and new jobs, and changing skills. However, he suggested, there will still be work to be done.
“We’re not going to run out of things to do,” he said. “We can eventually adapt to this kind of change. The deep human drivers don’t seem to be about to go anywhere.”
Interestingly, turning to talk of regulation, Altman seemed to offer a kind of flexibility in what society chooses to do.
“I get the impulse to say that any regulation is unacceptable,” he said, citing the potential for corporate regulatory capture. “If society decides that they don’t want to regulate AI at all, then we’ll just take our chances. I’ll accept the outcome of a democratic process. (However,) it seems, to me, good to have some voices saying ‘let’s not act out of fear, but proceed with some reasonable caution.’”
Asked about the potential for chaos, for example, in elections, he predicted that deepfakes, troll farms and other nefarious aspects of tech will continue to proliferate and evolve.
“I wish we were taking it seriously,” he said.
Turning toward the positives of this powerful technology, he noted that this is the first era in which computers are talking to us in our own languages, not in machine language or programming syntax.
“You can talk to a computer,” he said. “Between AI and people, you have the same kind of interface.”
That, he theorized, might lead to an ongoing cycle where tech always seems to be just taking off.
“You’ll always be able to use the tools to do amazing things,” he said. “We will just be able to accomplish more, to do more things – we’ll feel like this wave is about to crash over us…”
Pretty bullish!
But as I pointed out in my article contrasting Altman’s two talks at both MIT and Cambridge, in the end, he is always measured about AI, pointing out positives and negatives, the potential for good or ill, and the double-edged sword that AI represents for us all.