I’ve admired Bill Gates’ ability to analyze technology advances from a practical and scientific perspective. In his most recent blog post, Bill writes that AI is about to change how we will use computers in the future.
Bill summarizes the state of computers today when he writes that to do any task on a computer, we must tell our device which app to use. He writes that this will change over the next five years. We will tell our device what we want to do, and the software will respond personally because of its knowledge about our lives.
Software that responds to natural language and can complete many tasks based on its knowledge of the user is called an agent. Bill states that he has written about agents in the past (more than 20 years ago in fact), but only recently have they become more practical because of advances in AI. His description of what agents will do is worth quoting:
“An agent will be able to help you with all your activities if you want it to. 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. It will get your personal and work relationships, hobbies, preferences, and schedule. You’ll choose how and when it will step in to help with something or ask you to make a decision.”
The AI tools that most of us use today, Bill writes, are bots. They are limited to one app and step in when we ask them for help. They don’t get better or learn our preferences. Agents are smarter. They are proactive. They improve over time. They offer to provide what they think you need.
Bill adds that the most exciting impact of AI agents will be that “they will democratize services that are too expensive for most people.” The four areas he believes will be influenced the most are healthcare, education, productivity, and entertainment and shopping.
Agents Will Be the Platforms of the Future
Today’s apps are built on the familiar operating platforms of Android, iOS, and Windows. Bill states that agents will be the next platform.
People who want to create a new app or service won’t need to know how to write code or do graphic design. They will just tell their agent what they want. Bill adds that OpenAI’s release of its new GPTs service that allows anyone to build a customized version of ChatGPT is similar to what he envisions a future agent platform will enable.
Agents will replace search sites because they will be better at finding information and summarizing it for us writes Bill. Agents will also diminish the use of e-commerce sites because they’ll find the best price for us. Agents will replace spreadsheets, word processors, and other productivity apps.
Bill believes that there will be many agents created and because of their numbers, they will be inexpensive to use. There may be agents that are free because they are supported by advertising, but most will charge a fee.
Technical Challenges for Agents
Bill Gates lists several issues that need to be resolved before agents are developed. These are:
- No one has figured out what the data structure for an agent will look like.
- No one knows how many agents people will interact with.
- No one is sure how people will prefer to interact with their agent. Options include:
- There is no standard protocol for agents to talk to each other.
- Hallucinations need to be prevented.
- Biases need to be prevented.
- We need to make sure that agents don’t harm people.
Other Big Concerns
Resolving issues related to online privacy and security will become more urgent writes Bill. People will need to be able to decide what information the agent has access to so that data is only shared with people and companies chosen by each of us. No one wants to receive ads due to information they shared with their therapist, for example.
Bill writes “in the distant future, agents may even force humans to face profound questions about purpose. Imagine that agents become so good that everyone can have a high quality of life without working nearly as much. What would people do with their time? Would anyone still want to get an education when an agent has all the answers?”
How Should We Plan for Agents?
Bill Gates’ successes have afforded him the ability to think about big picture ideas like agents. His thoughts are valuable for those of us with less time to meet with futurists and technology experts.
Bill predicts that agents will come into relevancy sometime over the next five years. The three platforms that he states will be replaced by agents, Windows, iOS, and Android, were launched in 1985, 2007, and 2008. The latter two were different in that they were built specifically for mobile devices.
If an independent agent is developed as described by Bill, it’s likely to be an offshoot of OpenAI or another LLM supported by a deep pocketed tech company like Microsoft, Amazon, Google, Apple, or Facebook. It’s no secret that corporations have outpaced universities on the research and development spending on AI.
I haven’t had the time to digest the six principles that more than 140 higher education organizations collaborated to guide development of AI policies and practices at colleges and universities. The development and implementation of agents appear to align with these principles assuming some of Bill Gates’ concerns are addressed.
When I think about Bill Gates’ question “would anyone still want to get an education when an agent has all the answers?”, the answer is of course. It is ludicrous to believe that agents will control all information about the world and its citizens. Mankind has many problems to resolve. Technology has boosted the lives of those with access to it, but many nations are still behind.
I can afford a laptop and a phone that provide me with access to the internet and many apps and tools. I can also afford the $20 per month that OpenAI charges for access to ChatGPT Plus. Sadly, not everyone has access to those devices, apps, and tools.
It’s easy to observe that not everyone who can afford technology wants to embrace it as strongly as Bill Gates suggests we will embrace agents. Some of us might accept the loss of privacy and some of us may not. Some of us may choose to live a tech-enabled life, and some of us may choose to live “off the grid.”
Agents will undoubtedly be in our future. How soon is anyone’s guess. My recommendations to educators and leaders in education are to provide funding and time for teachers to experiment with the latest AI tools especially given that their students are experimenting with AI as well. People will continue to value and want an education to get ahead of those without an education. Teachers will enhance our ability to learn to maximize our thinking beyond querying an agent.