Stay informed with free updates
Simply sign up for the Artificial Intelligence myFT Digest, delivered straight to your inbox.
Ten years ago, after selling Siri to Apple, the developers of the voice-controlled AI assistant had another big idea.
With their new company, Viv, they set out to solve a persistent problem for smartphone users: how to accomplish everyday tasks without juggling multiple apps. Wouldn’t it be great if you didn’t have to switch between separate travel, hotel, and mapping apps when planning your vacation, and instead could rely on software that brings them all together?
Viv never achieved the breakthrough it had hoped for and was later incorporated into Bixby, Samsung’s AI assistant. But in technology, good ideas rarely die. The idea is just waiting for the underlying technology to advance to make it possible. As with many other things, that advancement came in the form of large-scale language models.
So essentially creating a new layer of digital plumbing between apps and websites doesn’t seem like the most appealing use of AI. But it could lead to important changes in how people use technology and shift the balance of power in the tech industry.
AI agents that act on behalf of users are the rage right now. Giving them the ability to operate across different apps, websites, and digital services can have far-reaching implications.
For example, AI startup Anthropic recently unveiled an AI system that interacts with a computer screen much like a human would. The demonstration, called “Using a Computer,” showed how the technology extracts information from a variety of sources to fill out online forms. This is something of a mundane, low-level task for many back-office workers.
The idea of using software programs to reproduce exactly what workers do on a computer screen has been around in other forms for years. Known as robotic process automation (RPA), this involves programming virtual “robots” to perform tasks across a variety of apps. The natural language capabilities of generative AI have given new life to this idea. Although Anthropic’s technology is designed to interact with computers in the same way humans do, the software is not yet adept at performing common operations on a computer screen, such as scrolling.
Services like this, which replace routine on-screen tasks for many office workers, could be the first real manifestation of generative AI. Matt Garman, head of Amazon Web Services, described his company’s latest effort to automatically coordinate work between groups of AI agents to complete more complex tasks as “powerful RPA.” I did.
Meanwhile, the closest thing to Viv’s idea of integrating apps into the consumer world comes from Apple. The feature, known as app intents, requires developers to make their apps work in conjunction with Apple’s AI, allowing software to work across apps without the user having to open the app.
The impact of this seemingly mundane idea can be far-reaching. When an AI assistant or agent automatically has access to the data and functionality it needs to complete a specific task, you open fewer apps, visit fewer websites, and use fewer digital services. Masu. In fact, this results in a person’s digital activity being concentrated in fewer places.
One of the consequences is that app developers have decided to make their apps one of the primary places to control users’ attention and continue to serve as a hub for completing tasks, rather than risk being relegated to ancillary status. There is a possibility that you will be in a hurry to continue to be.
Eventually, a small number of general-purpose AI assistants that function like automated super apps may dominate. If that were the case, the value could be sucked out of many independent apps as users would no longer open them or pay attention to them.
This poses a dilemma for app developers. Opening up services to big tech companies’ AI may mean losing direct contact with users. But any attempt to differentiate could lock them out of a new digital ecosystem that is likely to coalesce around AI agents.
Meanwhile, the biggest tech companies will be able to seamlessly integrate their AI agents with their apps, giving people even more reason to gravitate toward their technology.
This will create new challenges for antitrust regulators. Just as the biggest technology platforms are beginning to figure out how to direct users to their services (a practice known as prioritization), a whole new layer of technology is starting to connect technology users even more closely to Big Tech’s digital universe. There is a possibility that it will appear. .
richard.waters@ft.com
