It is becoming harder to figure out where we are with AI. At one end of the scale, AI seems to be the answer to, well everything. According to Demis Hassabis, CEO of DeepMind (now owned by Google), “Step one is to use it to figure out natural intelligence. Step two is to use it to solve absolutely everything else.”
This may sound fanciful, but then you hear that Elon Musk has launched his company, Neuralink, that will produce neural laces, an interface between the brain and the web that will allow the brain to keep up with artificial intelligence.
And just as you are thinking, “Well that’s Elon Musk, he’s somewhere between genius and mad, he would do something like that,” along comes Mark Zuckerberg.
Zuckerberg has joined the classes of mad futurists by announcing that he has a team that has been working on a direct brain interface. His goal is to be able, “eventually, to communicate using only your mind”. Oddly, we thought that was how we communicate now, but you can see where he is going. Just.
Back towards the realms of near sanity, we read that Telefonica has been looking at AI and will be using it to process anonymized customer and network data to better manage networks and customers. They will be rolling this this in Germany, Argentina and Chile. So not very close to home, then.
The air is thick with cynicism and skepticism.
Many – including ourselves – believe that most AI initiatives are little more than very, very fast look-up machines. And there is nothing wrong with very, very fast look-up, if it helps solve a problem or helps you do something better.
It just seems as if most of the companies that are into AI are not as good at it as they think they are. As Richard Windsor points out, Samsung’s delay in rolling out a lot of features of its supposedly AI-powered digital assistant Bixby just shows that Samsung is not very good at it.
Facebook, while its head is in the augmented reality of Mark Zuckerberg’s brain, is facing serious challenges with its inability to control content on the site. The sheer volume of the content dictates that AI be used to filter content. It is not working and Facebook will be sued quite soon, one would imagine.
Google is similarly in trouble, allowing big brands and government adverts to be placed next to racist, extremist or homophobic content. Same reason.
It seems, then, that the scope of AI news stretches from “behind the curve” to the “extremely fanciful”.
We also hate to point this out to Mr Z, but if he achieves his stated goal of eliminating screens from our lives and simply wearing contacts, glasses or some other direct brain interface, where will he put all those adverts that make up the vast majority of his revenue?