Companies frustrated with this type of AI are a bad thing

“This is very cool, but I can't actually make it work reliably enough to offer it to our customers.”

Reality check

By now, it seems clear that much of the hype around generative AI is overblown — if not a bubble bound to burst — and some companies that have invested in the technology are learning that the hard way.

It is difficult to overlook the shortcomings of technology. Large language models like ChatGPT are vulnerable to hallucinations and spreading misinformation. Both chatbots and AI image makers have been accused of ripping off writers and artists. In general, the devices used by artificial intelligence require huge amounts of energy, which leads to environmental destruction.

Perhaps most important of all, according to Gary Marcus, a cognitive scientist and leading AI researcher, companies are discovering that technology cannot be relied upon.

“It seems like almost everyone comes back with a report that's like, 'This is super cool, but I can't actually make it work reliably enough to deliver it to our customers,'” Marcus Tell Axios.

customers service

This is not hard to believe. A British company was forced to disable its chatbot after it started insulting customers and speaking inappropriately to their employers. A California car dealership had to do the same after a ChatGPT-powered car seller started offering $1 cars to buyers.

Even more depressingly, an airline had to pay damages after its chatbot lied to a grieving customer, telling them that if they bought a full-price ticket to their grandmother's funeral, they would be guaranteed a bereavement discount later.

“No one wants to build a product on a model that makes things,” said Roman Chowdhury, CEO of AI consulting firm Humane Intelligence. Axios. “The fundamental problem is that GenAI models are not information retrieval systems,” she added. “They crowdsource systems, without the ability to differentiate from the data they are trained on unless significant guardrails are put in place.”

Bubble problem

For these reasons and many more, some experts have warned that artificial intelligence is a bubble, as are cryptocurrency or dot-com startups. Expectations that artificial intelligence will become An industry worth more than a trillion dollars The coming decade certainly seems to remind us of the hubris that always comes with those booms before busts.

Meanwhile, from a technical standpoint, some doubt that the technology will advance fast enough to keep up with the hype, and may even go through a long period of stagnation. Investors who have poured billions of dollars into the industry in a very short period of time, expecting a lucrative turnaround, may not have the patience to hold out.

“It's easy to say, 'We're only a few months away,'” Marcus said. Axios. “I don't think we're in this particular situation. Not because I don't think AI is impossible or that AGI is impossible, but just because [this] “A certain technology has a lot of problems.”

More about artificial intelligence: VC criticizes Sam Altman, calling him “megalomaniac”