5 errors Sam Altman and Jony Ive must avoid to prevent their chatgpt AI device from going the way of the rabbit R1 and Humane Ai Pin
- Advertisement -
It is not often that two of the biggest names in technology come together to “completely reconsider what it means to use a computer”, but that is what happened as Sam Altman Openi Recently bought the IO design company from Jony Ive and announced that together they would make a number of new AI products.
Jony Ive is perhaps best known for designing the iPhone, and Sam Altman is the driving force behind ChatgptSo my mind immediately jumped on the idea of the world’s first chatgpt-driven smartphone, but it turns out that Ive and Altman have something else in mind. Something we may have seen before.
It was all revealed in a video of nine minutes that OpenAi released yesterday, called Sam and Jony introduce Io:
The video starts with Altman who says: “I think we have the chance here to think a little again what it means to use a computer”.
According to Io, Io is “formed with a mission to create a family of devices with which people can use AI to create all kinds of beautiful things”. The video also mentions a new prototype of IVEs that Altman has tested.
Altman calls it “the coolest piece of technology that the world has ever seen”, but no details about it, although instructions are littered during the video.
In one segment, for example, he talks about how cumbersome the process of using a laptop is because it has a screen.
It is difficult not to be wiped out in the reality deflation field that this technical gurus surrounds, but making an AI computer without a screen in essentially means something small that you wear or participate and communicate with the use of your voice, and that concept has so far more than one technical startup.
Of course I refer to the now legendary technical failures that are the humane AI pin and the rabbit R1.
Radiate me on scotty
The human AI -pin was in fact like the classic Rigid trait battle. It was right on your clothes and stayed there in the standby mode until you were working on it. It was ahead in many ways – you could ask questions (such as an AI -driven Siri or Alexa) and having it playing to play music. It also had a camera, so it could see what you looked at and answer questions about it. It can also shoot a laser on your palm that has staged a menu system with which you could communicate with the use of hand gestures.
The rabbit R1 was similar, except that it was designed to fit on your pocket, not worn on your chest.
They were both good ideas, and their time before their time, but good ideas are not automatically in good products. They were each poor in a number of fundamental ways.
Here are five classic mistakes that people have already made with AI devices that should avoid Ive and Altman:
1. Sometimes you need a screen
The human AI -pin was built around the idea of replacing your smartphone, but that would never work.
In theory, combining the power of advanced AI and the entire internet, in a way that means that you could still be fully present at the moment and not had been gone by intrusive and addictive telephone applications, a good idea. In practice it just didn’t work.
We live in a world where everyone already has a smartphone, so if you want to communicate with other people who all use screens and share visual information, then you have to use a screen at some point.
Or that means that the IO device can cooperate the existing screens in your life, or it can somehow be attached to a screen, I don’t know, but remember that screens are still essential for some things that will be the key to its success.
2. It cannot be too expensive
This becomes the key to the device acceptance. A major criticism of the AI devices released so far is that they were too expensive. We already have an expensive smartphone and a data plan to go with it, don’t expect that we can afford an equally priced second device and a second data plan every month.
Jony Ive gave us the iPhone and the MacBook Pro – devices that defined technology for a generation, but those products have always had a premium price tag.
The man can be a genius, but if this product wants to succeed, it needs a price that is logical, not one that alienates his target group.
3. The voice interaction must be flawless
Ai -speech modes, in particular the advanced speech mode of Chatgpt and Gemini Live, have taken a long way in recent months. They feel much more natural to use, as if you have a real conversation with a friend.
Delivering that experience with the new IO device is the key to its success. If a conversation with the AI is bent in any way or if there are long breaks after you have told something to do – a problem that has plagued earlier AI devices – then it will not start.
The whole experience to use it must be as frictional as possible.
4. It can’t hallucinate at all
AI tends to come up with things known as hallucinating, and it is a problem that seems to be There get worse, the smarter the AI gets.
According to its own reports, the new O3 model of Chatgpt has included hallucinations in a third of a benchmark test with public figures, which is double the error percentage of the earlier O1 model. The compact O4-mini model still performed poorly, hallucinating at 48% of comparable tasks.
A theory that explains why AI gets worse in hallucinating is that the deeper the reasoning method goes, the greater the chance that it must mess up. The older, simpler models clung to predictions with a higher confidence, in contrast to today’s complex models. When there are several paths to consider, the model has a greater chance of improvising and being wrong.
Whatever the problem is, the Human AI -pin, which could use any AI model, certainly hallucinated, and there is no trace of the problem in every product released by IO if it wants to be trusted.
5. It must work without an internet connection
This will be a difficult thing to certainly implement, but it requires a second data plan only for the device was one of the biggest problems with the human AI -pin. It was useless without access to the internet. Perhaps the new IO devices can display the internet connection from your phone, but Ideally, they would also be able to process on the device, which means that they can continue to function, even in places where there is no data reporting.
Back on May 7, Apple Exec Eddy Cue slide that “you may not need an iPhone in 10 years”, while you give a witness to the Google Search Antitrust Remedies Trial.
It is clear that the time is ripe for a kind of AI device, and the possibilities to reconsider a human and computer interaction are endless, but as we have seen, others have gone on this road earlier, and so far it has not ended well. Let’s hope what Io will be released in 2026 (the promised delivery date for his AI -products family) can discover this trend because no one wants to see another doomed AI device on the market.
Maybe you like it too
- Advertisement -
