Forget photo manipulation and art – the real potential of generative AI helps us with things we don’t know and can’t see
You couldn’t swing a virtual cat without hitting an MWC 2024 exhibitor talking about generative AI; from Samsung and Google touting AI-powered photo editing and search, to Qualcomm wanting to put AI in cars and Humane wanting to make AI portable.
Sure, a lot of the technology on display was cool, but I don’t think the real game-changing potential of generative AI is in the obvious stuff. An experienced photographer and editor will almost certainly want to make their own edits – or simply take a photo – rather than relying on AI manipulation. And writers like me would rather type out an article than try to understand something written by ChatGPT.
Instead, based on the various talks and roundtables I’ve participated in at MWC, I think AI will be most useful when it helps people do things they can’t easily do, learn, or even discover.
A simple example of this is how Qualcomm continues to push AI use in its modem chips, with the latest Snapdragon X80 leaning on a dedicated AI tensor accelerator to boost connectivity speeds, optimize power consumption, and generally improve latency and coverage. These are certainly not the sexiest tech products, but in consumer terms this means more effective Wi-Fi and mobile internet connectivity with lower power consumption in smartphones and other 5G devices; Since Qualcomm supplies modem chips for all major phone brands, you can expect the Snapdragon X80 to show up in future Galaxy S-series phones and the iPhone 16.
Similarly, the Snapdragon X80 can be used in so-called Internet of Things (IoT) devices and applications, which could improve the performance of smart sensors and networks in smart cities. You may not see or notice this, but it could be, for example, smart traffic monitoring that is fed back to apps such as Google Maps. Or just fast and reliable mmWave 5G connectivity; basically things that make everyday life easier while you function in the background.
Live translation, already a generative AI feature on the Google Pixel 8 series and the Samsung Galaxy S24 family, is a clear example of AI being used to help us deal with things we don’t know, although I have yet to trying out.
Then there’s the use of generative AI in cars, which is especially interesting because it allows people to talk to their cars. Qualcomm showed off its Snapdragon Digital Chassis platform where people can ask a car-based generative AI questions ranging from where to eat nearby to what a specific warning light on the dashboard means.
Getting recommendations isn’t anything special since Google Maps already does that and it’s built into Android Auto. But getting information about what might be wrong with your car without the driver having to dig through a manual or go to a garage could be a fantastic next step for automotive technology. Additionally, generative AI can supercharge voice commands and make it a lot easier for drivers to activate different settings and options without taking their eyes off the road or their hands off the wheel.
Generative AI in robots also caught my attention, with the almost creepy Ameca robot showing up at MWC to surprise attendees who threw all sorts of questions at the humanoid bot.
Here, AI isn’t necessarily used to help with one specific thing, but rather to power a smart machine that can be useful in all kinds of tasks. I saw Ameca being used to provide 24-hour concierge service in hotels and airports – not to replace people, but to fill the gaps between shift changes and during quiet times.
Or perhaps a next-generation Ameca could be used as a companion for those in social care or in need of care, but still want their independence. At a time when there is a shortage of social workers (at least in Britain), an AI-powered robot could at least help with staff shortages.
I have no doubt that we will continue to hear about AI in consumer technology, and I would bet a good amount of money that a next-generation iPhone will have some generative AI tools. But I’d like to see how generative AI can quietly improve lives with better efficiency of services, easier ways to surface difficult information and help people, rather than simply taking pictures of dogs on the moon or overly pristine images to produce.