What is happening with the new startups what they want to develop AI Gadgets and why do they fail?
Today we have to do a bit of criticism - hopefully constructive - in order to try to explain the phenomenon that we have been experiencing since the beginning of this year 2024 with the devices powered by artificial intelligence.
And one after another, as if it were some kind of unwritten rule, these gadgets are failing one after another, they are not reaching the general public and they also do not seem to live up to what users expect. Because? Let's see it.
The novelty of AI
As I said at the beginning, we are talking about a whole new worldWe can say that the use of artificial intelligence at a general level - although there are hardly any users as a percentage of the population - did not appear until just 2 years ago.
And that was when OpenAI developed version of ChatGPT that the world knew as a paradigm shift with respect to AI technology and opened the doors for anyone to access it.
From there and almost out of nowhere, they emerged hundreds of AIs that promised to help us with different problems from day to day, from text, images or video to music, science or medicine. It seemed like we went from 0 to 100 in one year.
And this is one of the first points related to the AI Gadgets, and it is that all of them work to a greater or lesser extent with these artificial intelligences which, although they are advancing by leaps and bounds, are still based on embryonic technology and are taking their first steps.
This causes AI Gadgets to have an internal dependency on a technology that is not defined, refined and corrected, with which the failures or hallucinations of the AIs have a negative impact on the devices irremediably.
Unfulfilled expectations
Perhaps one of the key points of this whole matter is precisely the mistake that both users and companies have made in generate expectations that they cannot meet and that we as clients should not have believed.
Let's put the Rabbit r1 case, a device powered by artificial intelligence that was going to be able to do for us anything that required interaction with a website or an app on a mobile phone.
In it CES 2024 we could see that this device You could order flights, organize trips and ultimately we could give you control of any website just as we would handle it. We could even teach you how to do it.
To this day, it still has problems with the apps that come by default —Spotify, Uber or DoorDash— and they do not work well to the point that they are impossible to handle consistently. It may work in the future, but it's more of a leap of faith.
AND what to say about Humane AI Pin. More of the same. A battery that does not last, that heats up and to top it off threatens to burn. In addition to a laser that cannot be seen in sunlight and whose handling is rather a torture…
The form factor of AI Gadgets
This is one of the problems that will be most difficult to solve. Whether we like it more or less, we are completely dependent on mobile phones and one of the efforts of the creators of AI devices is to change it.
And it's not easy at all. On the one hand there are wearables, such as the AI Pin, the Ray-Ban Meta, whose effort is to have a gadget that we can wear and that is not invasive on a daily basis or the Limitless Pendant that I will talk about soon.
On the other hand, we can count on devices such as the Rabbit r1 or the 01 Light by Open Interpreter, a kind of plastic spheroid that we can ask to execute actions on our computer, giving us all control.
But these form factors don't work, or at least not in the short term. We are too involved with mobile telephony and the fact of carrying a second device or having something hanging on your chest that looks like a camera is not being liked.
I imagine that like everything it will be a matter of time, but it seems to me that there is a long way so that all these devices end up adapting to everyday use.
Materials of dubious quality
We're also seeing a pretty strange effect on these devices. We have been searching for many years in the smartphones the best screens, processors, batteries or cameras, and it is taking its toll on new AI devices.
He hardware It's basic. The Rabbit r1 could pass for a mid-range mobile phone from years ago. The Pendant or the 01 Light don't even have a screen and are a plastic cover, and apart from the defective Humane batteries, the laser doesn't do much of anything either.
They want to be able to give customers the best device at the lowest price, but they are actually choosing to adjust the BOM —Bill Of Materials or production cost— so much that sometimes it feels like they are falling short, and that penalizes.
It's too early for AI hardware
This last part is as good as it is bad. We are like when the first mobile phones appeared, the first smartphones or the first generations of consoles. It's the beginning, and that's exciting.
I'm not saying that it's not their time, but that those who We are pioneers in the acquisition of this type of devices We have to be aware that we are at the beginning of the path, like when mobile technology was incipient.
We have to be patient, know how to choose and get used to the idea that there is still a long way to go. Let's see how far we go.
