As Generative artificial intelligence has begun to take possession of objects, there comes the emergence of AI Smartphones. Some wearable gadgets such as the R1 Rabbit and the matchbox-sized AI Pin that can be controlled entirely by voice have already appeared.
For some, they will take the place of cell phones but it will take time. In the meantime, smartphone manufacturers have converted to AI almost en masse. The new multi-year alliance between Google and Samsung announced a few days ago has shown that AI phones are a reality.
The new Samsung Galaxy S24 listens, records, translates, suggests how to manipulate images, and helps you with video editing. It does so using both Google’s APIs – Gemini and the Gen AI services.
The Google Pixels AIs are also ones to be excited about. The Chinese companies Honor, Motorola, Oppo, and Xiaomi are also moving in this direction both through agreements and by working on proprietary chatbots. It means more generative AIs on smartphone devices.
Magic of the S24 Series AI Phones
The AI Smartphones promise to bring on board the magic of the AI gen that we have known with ChatGpt and image generation software like Midjourney. It means visual searches with Google’s Circle and Search feature, deleting photo elements, video editing suggestions, and numerous incremental innovations in computational photography.
The most interesting feature of the S24 is a real-time translator. It means voice and text translations of calls in real-time bi-directionally within the native “phone” app. You can call a Korean restaurant speaking your language and let AI do the translation in real time.
These operations require a lot of computing power. Data transmission consumes a large amount of energy. Sending sensitive data to AI servers still presents some unknowns, there are doubts related to privacy, and the regulatory framework apart from the European AI Act is still fluid. For users (and not only) knowing that they have their data inside their device is reassuring.
At this moment we seem to understand the challenge for consumer electronics manufacturers is to bring generative artificial intelligence in-house as much as possible and rely on the big AI companies.
The iPhone will increasingly be the result of agreements between hardware manufacturers and AI providers. Samsung, for example, is allied with Google on smartphones with Gemini but also has collaborations with Microsoft CoPilot’s AI Gen on computers.
We expect that Apple will present its own AI Gen on device, proprietary so to speak, leaving developers the possibility of offering artificial intelligence services with the big names in AI. A new non-trivial competitive geography is emerging where those who can will try to work on their own AI on devices.
Samsung Research, the research and development division in November announced the in-house development of a generative artificial intelligence model called Samsung Gauss, named after the mathematician Carl Friedrich Gauss.
It consists of three models: Samsung Gauss Language, Samsung Gauss Code, and Samsung Gauss Image. Galaxy AI uses some aspects of this model. But the majority to date is managed by Google.
Related: Midjourney API
AI Smartphones – Final Thought
Looking ahead, users will need to know what is processed on the cloud by the AI and what remains on the device.
However, what we still don’t know is how much AI services in smartphone format will cost. That is, what will be the business model of AI on device?
In the case of the Koreans, we officially know that the Galaxy AI features will be available for free via Samsung Galaxy devices that support them until at least December 31, 2025.
This means that if tomorrow ChatGpt, Copilot, or Gemini start charging for their AI services we will understand better how much the new economy of AI Gen is worth. For now, we have a date. The end of 2025.