Meanwhile, in invisibly, AI is already transforming education, commerce, and the workplace. A friend recently told me about a large IT company where he works. The company had a long-established protocol for launching major initiatives that included designing the solution, coding the product, and designing the rollout. It took months to go from concept to execution. But he recently saw a demonstration of cutting-edge AI applied to a typical software project. “Everything that took months happened in just a few hours,” he says. “So I resonated with your column. A lot of the companies around us are now walking corpses.” It’s no wonder people feel scared.
A big part of the outrage over AI is distrust of the companies that develop and promote it. Coincidentally, I was having breakfast this week with Ali Farhadi, CEO of the Allen Institute for AI, a nonprofit research institute. He is 100% convinced that the AI hype is justified, but he understands the feelings of those who don’t buy into it. That’s because companies that try to dominate the AI field are looked upon with suspicion by the public, Farhadi says. “AI has been treated as a kind of black box that nobody knows about, and it’s so expensive that only four companies can run it,” Farhadi says. The fact that AI developers are moving fast further fuels distrust. “We collectively don’t understand this, and we’re deploying it,” Farhadi says. “I’m not against it, but we have to assume that these systems will behave in unpredictable ways and that people will react to that.” Farhadi, an advocate of open source AI, says that at the very least, big companies should be public about what materials they use to train their models.
Further complicating the issue, many people involved in building AI have also pledged to commit to creating AGI. While many leading researchers believe this would be a boon to humanity (it’s the founding principle of OpenAI), they haven’t publicly touted it. “People are unhappy with the idea that this AGI is going to happen tomorrow, or a year from now, or six months from now,” says Farhadi, who is no fan of the concept. He says AGI is not a scientific term, but a vague concept that will kill AI adoption. “In my lab, if a student uses those three letters, it just delays their graduation by six months,” he says.
Personally, I’m skeptical about the AGI issue. I don’t think we’re on the brink of AGI, but I don’t know what’s going to happen in the long term. When I talk to people at the forefront of AI, I find they don’t know either.
Some things are clear to me, and I think they will eventually become clear to everyone, even to those who badmouth me for X. AI will become more and more powerful. People will find ways to use AI to make their work and personal lives easier. Many people will also lose their jobs and entire companies will be thrown into disarray. Even if the AI boom might create new jobs and companies, that will be of small consolation, because some of those who lost their jobs will still be standing in unemployment lines or working cashiers at Walmart. In the meantime, everyone in the AI world, including columnists like me, should understand why people are so outraged and respect their legitimate grievances.
