The era of artificial intelligence has only just begun. We are at such an early stage of this technology that, despite the fears (justified or not) it arouses, regulations have not yet even been established to try to mitigate its risks. Just now world leaders are beginning to propose control proposals. But in research laboratories around the world, technology companies don't stop working and making plans for how we are going to use it. The computer will continue to be one of the most common access devices, but the mobile phone, which already governs many other areas of our lives, will be the main center from which we will use AI.
The Tech World conference organized annually by the multinational computer company Lenovo in Austin (Texas) has been dedicated in this year's edition to – of course – artificial intelligence under the title of AI for everyone. The main leaders of some of the most important technology companies in the world attended the call, sometimes virtually and other times on stage: Satya Nadella (Microsoft), Jen-Hsun Huang (Nvidia), Cristiano Amon (Qualcomm), Lisa Su (AMD) or the host, Yang Yuanqing (Lenovo).
Companies have multiple intertwined plans to cooperate on applying AI across the board, from the business to the home. One of the most striking concepts of the proposals presented was that of the AI twin. In principle, it is somewhat disturbing that a company can articulate a kind of replica of a person in the form of an artificial intelligence model. Although the reality of the proposal is not as worrying as it may seem at first glance, legislators in Europe and the United States preparing regulatory legislation should take note.
The AI twin proposal, with which we would communicate via mobile phone, is that of an artificial intelligence that is closer than the current language models, generic and disconnected from the individual reality of each person. An offline example was shown on stage, in airplane mode.
If a user asks an AI bot like ChatGPT to plan a travel itinerary to a foreign country, the response is standard and depersonalized. On the other hand, the AI twin that lives inside the mobile phone itself bases its entire response on the tastes and circumstances of the user, of whom he knows multiple details, since he has collected a huge amount of data over time. He thus decides which airports, activities, hotels and restaurants the person prefers.
It may sound terrifying that your virtual twin knows so much about yourself. Lenovo officials explained that, thanks to the advances in chips that are being developed in recent years, mobile phones will have enormous processing capacity capable of handling AI within each phone, without their data going outside and preserving privacy. of its user, who will decide exactly what aspects of their life they want them to learn.
Having private AI models, which will then learn within each device from the data of each individual user, has some advantages in addition to greater privacy. Large language models like OpenAI's GPT-4 run on supercomputers and require a lot of power to run. The massive use of artificial intelligence on devices, not just mobile phones, has a sustainability effect.
It is precisely at that point that another of the great proposals of these technologies for consumers is born. Since the 70s of the 20th century we have used PC (personal computer). In a few years, which the organizers estimate in two or a little more, we will have AI PC (artificial intelligence PC), that is, personal computers with internal AI, like telephones.
Lenovo showed advances in Austin such as Project Chronos, a system that consists of applying artificial intelligence to capture a person's movements to simultaneously translate sign language, which is different in each country. The host company also showed the central processing unit (CPU) that will equip Nvidia's Drive Thor platform, a powerful computer for vehicles.
But the most astonishing application of AI in both consumer and enterprise technology came with a computer monitor. The ThinkVision 27 3D allows you to view content in three dimensions without glasses. A camera searches the user's eyes. The AI interprets how to send the images to each eye to create a stereoscopic effect. Amazing. In a few years, it can be applied to five users at the same time.
