Will AI cause or solve more problems?

Five years or so seems to be the sweet spot for people and companies working on generative AI. It’s a stretch of time lots of folks, like Spectrum Labs’ Justin Davis have achieved. Davis joined GamesBeat’s own Dean Takahashi to talk about generative AI and whether communities are ready for it during our recent GamesBeat Summit.

That five year timeframe is popping up all over the place. A lot of the big names in generative AI technologies have crossed it, but for the most part have kept it quiet. It takes time to create working technology, and the AI space is no different. 

As consumers we only see the end result. We see something like ChatGPT get released, and multiple versions and revisions quickly follow. It creates the illusion that these things are normally quick to develop, but the reality is that it usually takes time.

That time spent has an additional effect. It lets creators think about how the technology will end up being used by the public at large. AI developers have to consider all the angles, including bad-faith uses which have already begun to pop up.

“I think that’s just a sign of what’s to come,” said Davis, about an AI generated hoax causing a drop in the market. “We’ve been talking about deepfakes and all sorts of threats coming from generative AI for many years now. It was just another reminder of the risk and the threats that happen when you put powerful technology in the hands of folks that can actually do some real world harm.”

The technology is there

Despite a handful of examples similar to that, the truth is that most people don’t have a solid grasp on AI yet. But with software like ChatGPT available in the wild, things like Unity’s Project Barracuda moving toward mainstream devices it’s only a matter of time.

The fact of the matter is that it’s almost less about whether or not communities are ready for generative AI. The technology is there, and it’s improving every day. As companies begin to use it to build out video game worlds, or use it for moderation purposes, communities won’t have much choice but to get ready.

For now, though, there’s still a necessary human element in play. AI can do a lot already, but it can’t do everything. Yet.

“You come up with a hate speech or threat policy that covers all the use cases that the AI can support,” said Davis. “But what about the edges? What about the things that aren’t covered by that policy? That’s where you need humans to come in.”

Originally appeared on: TheSpuzz