Computers today require the user to lead the machine. That’s also true in the online advertising industry, despite the amount of automation we have seen built into platforms in the past few years.

Tomorrow, computers will run on agents.

Despite the headline of this post, Bill Gates has a unique view. He often shares his thoughts in a blog about the future of computing. In a recent post about AI, he wrote: “in the computing industry, we talk about platforms— the technologies that apps and services are built on. Android, iOS, and Windows are all platforms. Agents will be the next platform.”

Android, iOS, and Windows are all platforms. On a device, the user queries a response in everyday conversational language, and the machine will determine which software to use and how to get the best result.

Within five years, the idea of an app will seem as outdated as a record player or rotary phone.

“An agent will be able to help you with all your activities if you want it to,” Gates wrote. “With permission to follow your online interactions and real-world locations, it will develop a powerful understanding of the people, places, and activities you engage in.”

The agent will have access to users’ personal and work relationships, hobbies, preferences, and schedules. The user will choose how and when it steps in to help with something or to ask the person to make a decision.

Microsoft is making Copilot part of Word, Excel, Outlook, and other services, while Google is doing similar things with Assistant with Bard and its productivity tools.

Gates is a software guy. If he would have invented AI, the world might have already had what he calls “agents.”

Elon Musk has also mentioned in the past that he wants to create one app that will run everything — his reasoning for changing the name of Twitter to X.

The biggest technical challenge is “nobody has figured out yet what the data structure for an agent will look like,” says Gates. To create personal agents, we need a new type of database that can capture all the nuances of your interests and relationships and quickly recall the information while maintaining your privacy.”

Gates points to new ways of storing information, such as vector databases, that may be better for storing data generated by machine-learning models, but the complexities of building a system with agents remain a challenge.

Will AI dumb down users — similar to the way people thought Google Search would make us dumb?

Gates explains that to create a new app or service, the user will not need to know how to write code or do graphic design. People will just tell your agent what they want. It will be able to write the code, design the look and feel of the app, create a logo, and publish the app to an online store. OpenAI’s launch of GPTs for small online businesses provide a look into future where non-developers create and share assistants.

Microsoft, the company Gates founded with Paul Allen, invested about $13 billion in OpenAI’s technology including ChatGPT, which will celebrate its first birthday on November 30. (Microsoft owns 49% of OpenAI)

On the day it was released, ChatGPT attracted 153,000 visits to chat.openai.com, and by the end of the first week, it had reached 15.5 million visits, growing to 58 million in week two, according to Similarweb.

Then came the release of DALL-E 2, which generates usable images based on a text prompt. These include usable images, and it also has the ability to expand and build on — as well as manipulate — images based on words the AI technology understands.

ChatGPT continues to grow — reaching 1.8 billion monthly visits by May 2023.

Similarweb estimates that parent company OpenAI’s website has become one of the most-visited domains in the world, breaking into the top 100 earlier this year and now on par with sites like LinkedIn, Netflix, and Reddit.

Gates did not invent AI, but he writes a lot about the power of agents — the models that will combine and run software without the user telling the machines what to do.

How did we get here? A series of events drove humanity to create AI. Alan Turing, a young British polymath, explored the mathematical possibility of artificial intelligence, suggesting that humans should use available information as well as reason to solve problems and make decisions. This logical framework became the backbone of his 1950 paper, Computing Machinery and Intelligence, in which he discussed how to build intelligent machines and how to test their intelligence.

Before 1949, computers lacked intelligence. They could store information, but could only execute commands. Computing was extremely expensive. Leasing a computer ran up to $200,000 a month in the early 1950s.

A proof of concept emerged five years later. Allen Newell, Cliff Shaw, and Herbert Simon’ created the Logic Theorist, a program designed to mimic the problem solving skills of a human funded by Research and Development (RAND) Corporation, according to a Harvard publication. Many consider this to be the first AI program. It was presented at the Dartmouth Summer Research Project on Artificial Intelligence (DSRPAI) hosted by John McCarthy and Marvin Minsky in 1956.

There has been a mountain of obstacles along the way. Fast forward to the 2020s, when great minds across the advertising industry sunk hooks into ways that AI could create, serve, bid a better ad to more responsive consumers searching for information or wanting to buy products and services.