AI technology is expected to disrupt 40% of all jobs by the end of the decade but its impact on smartphones could be even more widespread.
More than half the world’s phones are expected to use artificial intelligence technology within as little as four years, according to technology market analyst firm Canalys.
Electronic tech giant Samsung this week committed to putting AI features inside 200 million smart devices by the end of the year.
But technology experts and academics warn the rapidly evolving technology comes with unaddressed risks, including questions about consumer privacy, data security and transparency, and without regulation and education, users may have to rely on tech firms to do the right thing.
The forecasts come as Samsung unveiled AI features to roll out inside future smartphones and wearable devices at its Paris event, including tools to compose email and text messages, to recreate photos, and to summarise, transcribe and translate text.
Samsung Electronics president TM Roh says he considers generative AI technology so impactful that he made it the focus of the company’s mobile division.
“Internally last year, I proclaimed to the entire (mobile) business that we are no longer a smartphone business, we are an AI company, so we’re going to focus all efforts around AI,” he told AAP.
“With the advancement of AI technology, the industry as a whole has been influenced.”
But Samsung is not alone in its desire to put generative AI tools inside our pockets.
Google will hold a launch event on August 13 that promises to show off “the best of Google AI, Android software and the Pixel portfolio of devices,” while Apple unveiled its own AI software and a partnership with ChatGPT maker OpenAI in June.
Apple’s AI features, such as a smarter Siri voice assistant and writing tools, are expected to launch on devices in September.
All three major smartphone manufacturers say their AI features will be smarter, more useful, and more personalised to users.
For Samsung Galaxy AI that means analysing a text message conversation to suggest personalised responses, while for Apple that will mean scanning and prioritising a user’s most time-sensitive notifications.
Semiconductor maker and software creator Qualcomm senior vice-president Don McGuire says using generative AI technology proactively, rather than asking it to respond to prompts, is what will make it invaluable to smartphone users.
“In order for AI to be really useful for people … AI has to become more suggestive and anticipatory,” he said.
“That’s really where we’re headed.”
But OECD digital economy policy division policy analyst Lucia Russo says the fast-evolving technology comes with growing risks, including questions about when and where consumer data is shared, its role in spreading disinformation, and potential for bias.
Regulations to limit those risks, she says, are necessary, but tricky, as they must be able to evolve with AI technology and cross international borders.
“In the same way you have tech leaders coming together to ensure they make the best and most optimal user experience for products, it’s important for regulation at all levels to operate on mutual understanding,” she said.
“We need to balance the potential for innovation with the risks that this technology holds.”
The biggest consumer concerns about AI, according to Canalys, are privacy and data security, a lack of understanding and transparency about the technology, and the potential for extra cost to access AI tools.
Samsung and Apple have sought to manage those risks by offering consumers the option to process AI tasks on devices rather than over the internet.
Roh says 55% of AI use in Samsung phones, excluding internet searches, currently happens on the devices themselves, which is “higher than expected”.
But Swinburne University bioethicist Dr Evie Kendal predicts many smartphone users will not seek out their phone’s AI controls to ensure their data is not uploaded to the cloud, and many will not know how to distinguish where their data goes.
“People who are concerned about AI will be willing to do the extra work and go into the settings,” she said.
“But part of the problem with some new technologies is they get rolled out in a way that normalises them and makes them almost invisible and so people may not know that their information is being used in different ways.”
Education about generative AI tools, what they do, how they use data, and how to control your personal information should be an “immediate focus” for policymakers, Dr Kendal says, as well as consumers keen to try out the tech tools.
“You need to find out what you’re comfortable with, if you know how your data is being stored and shared, who has access to your personal information and make sure you’re comfortable with security of the device you’re buying,” she said.
“Beyond that, it’s a matter of being as well informed as possible when you make these purchasing decisions.”
The reporter travelled to France as a guest of Samsung Australia.
This article was first published by AAP.
Never miss a story: sign up to SmartCompany’s free daily newsletter and find our best stories on LinkedIn.