Why are they pushing AI (artificial intelligence) so hard?

If somebody has a factual answer, I’d welcome it, but I put this in IMHO because I am interested in speculation, too.

It seems that just in the last year, everybody (Google, Apple, Adobe, Outlook, my workplace) have really started pushing AI hard. There’s been something of a push for 30 years, at least, but what I don’t understand is why.

I’m particularly interested in the functions it pushes that I don’t want: writing correction (grammar / spelling / punctuation, but also finishing phrases and sentences for me); summarizing or highlighting text. These are things I do better on my own, and I would rather do on my own.

Other people can do what they like. What I don’t understand is why Google, Apple, etc. want me to use these. What are they getting out of constantly irritating me with unwanted, and often unwarranted, correction? Why do they make it so difficult to avoid or disable?

I presume the answer is money, but I don’t quite see how.

Employment insecurity.

Money-centered oligarchs want to drive down people’s expectations of having good well-paying jobs. People worried about being fired and replaced by AI will do more work for less money. So Wall Street is all-in on the new paradigm because it’ll cut the legs out from under the workforce.

I haven’t seen anything convincing to make me think there’s a more compelling reason than this.

There is a huge investment; they want to get to a point that it pays off.

Businesses are apparently slower to rush in than consumers are. The strategy, I’d WAG, is that if workers are using it anyway then the company may implement paid versions that actually leverage the tool more effectively in the workplace.

And it’s just the shiny new thing. Like in the early 00’s when they were jamming wi-fi and touchscreens onto everything from cars to refrigerators and coffee machines.

The Tech CEOs have invested everything into this and desperately need to make us like it so they get a return on all of the time and money put into it. They are so pot committed on this that they NEED it to work. The problem is every other recent “new thing” has failed, such as the metaverse, augmented reality glasses, etc. The investment is not just among the Tech CEOs either, this failing would cost a lot of people a lot of money. So it is being pushed on us from every direction.

They are pushing AI because they think they can make more money doing it. It’s as simple as that. Why would they need any other motivation? The early bird gets the worm.

Yes, but HOW do they make money?

I’ve stopped using Adobe to read pdfs because the AI pop-up is so annoying. I have uninstalled Cortana from my PC. Lots of other people are doing this, so objectively, they are losing a small number of users because of this; are they gaining any?

In 10 years, everyone with a cell phone or with a PC on the Internet will be using AI daily whether they know it or not. Wouldn’t you want to get in on the ground floor of that?

To get a return on the money they put in, for sure.
But there are more reasons. One is for competitiveness. If product X says it has AI, competitive product Y better have it also. And they have to worry about losing market share to new product Z.
Another reason is that AI, unlike much of tech, is very easy to sell. One of the Adobe ads has a business owner hit “write a newsletter” which it does and she copies and pastes it into an email apparently without reading it. (Way bad idea, but easy to grasp.)
So it is not making money from it, it is fear of missing out.

If you want to use AI in a significant way, you have to pay for the access and use. You might be able to make a few queries for free, but you’ll have to pay if you want to use it on a daily basis. These AI companies have invested billions in development and data centers to provide AI services. All the ads you see are them trying to get users to sign up to use their services. One reason for the ads are to try to get businesses which have already decided to use AI to use that tech company’s AI service. Another reason for the ads is to show people ways in which they can use AI so that they incorporate more AI services into their daily lives and ask their companies to sign up for more AI services.

I think @Mangetout has just made a video about that. I believe it is even worse than he states.

Not just make money, but save money. Why pay a human CEO an exorbitant salary when you can hire an AI CEO?

A useful tool for understanding where a technology is the Gartner Hype Cycle, which does not measure the maturity or utility of a technology, but the level of “buzz” round it. The 2024 Gartner Hype Cycle is shown below. The “pushing it hard” portion is at or near the top of peak hype. The mainstream commercialization is at the Plateau of Productivity. But if a company waits to jump in until it is commercialized, they risk losing out on significant market share, so companies try to get leadership early during the hype and ride that to full commercial value.

What’s changed in recent years is the utility (potential or actual) of generative AI (LLMs, ChatBots, etc.). Enable by the maturation of transformer technology, the breadth of addressable markets has exploded. So, despite the fact that much of generative AI is still more hype than practicality, companies are “pushing it hard”.

AI companies definitely charge businesses for using them. So, if they can get a bunch of regular folks to act as trainers and correctors for free, I can see why they would push it on consumers.

Years ago, Google provided a free Yellow Pages phone number service and used it to successfully train its voice model.

If I was the CEO who has invested huge in AI and I was seeing the numbers in the article below, you could bet I’d be promoting hard.

The other big fear? You’ve got to get the customers using your product so that switching to a much less expensive and only slight less cutting edge one is too much effort.

See the market reaction today to the release of China’s Deep Seek -

https://www.reuters.com/technology/chinas-deepseek-sets-off-ai-market-rout-2025-01-27/

It could be as simple as the premise of the movie “Robocop”: Omni Consumer Products was dependent on police to maintain law and order keep things from getting even worse in Detroit; but that meant dealing with salary issues, unions, pensions and health care, etc. If only there were a way to replace personnel with equipment…

LLMs and Image generators originated in academics: scientists wondering if they can get a computer to do x. When the answer turned out to be “yes”, of course people came along who wonder if they can make money off a computer doing x.

I get the green mumbles every time I see people use “AI” as an all-encompassing term.

There are lots and lots of AIs, for lots and lots of purposes. AIs are already indispensable work tools in the sciences, and are steadily advancing through all other industries.

The only way to think of AIs is to think of them as the modern equivalent of “computers.” Back in the 1980s and 1990s, people were throwing out wild claims about how computers would be everywhere and do everything for everyone. Lots of people loudly scoffed. Today, of course, computers are so ubiquitous only the most experienced expert could explain how they are used. Would anybody at the launch of the Apple 2 believe that 2024 cars would contain 1000 computers (using the term broadly for chips that run or control stuff as is now common)?

Now, take the ubiquity of computers and imagine increasing productivity by having them do, not just help, a lot of the work. Much of the prospective work is shifted over from the predictions of how robots were going to be doing everything. That never happened. But we have AI that works well for many things, even if some public utterances are defective. Chatbots are mere public relations nonsense, though. Besides we’re in the early, early days. AI will inprove greatly.

Chinese AI firm DeepSeek has emerged as a potential challenger to U.S. AI companies, demonstrating breakthrough models that claim to offer performance comparable to leading offerings at a fraction of the cost.

Nvidia stock dropped 16%. Doubts are appearing about whether these claims are true, but if not today then next year or in 10 years.

If you have a chance to get in at the floor of Apple and Microsoft wouldn’t you drop everything to do so? AI, in some future flavor that probably won’t even be called AI, is going to be in everything everywhere. We’re talking trillions, maybe quadrillions of dollars changing hands. That’s why there’s a push.

I’ve talked just about the business side. Computers were supposed to destroy jobs. (Robots, RIP, were supposed to destroy jobs.) AI is now supposed to destroy jobs. That’s another reason why the push is so strong. AI appears to be much more likely to do so than computers. We don’t know yet what AI-powered programs that nobody has yet thought of will help a bevy of new jobs to appear, however. The odds are not good but the unforeseen future is by definition unknown.

Thanks - yeah, it really does look like Microsoft had to show the shareholders that everyone loves Copilot, so they just made the decision on behalf of 365 users that they want to upgrade to a package that has it.

Yes, I know. This is why I tried to be specific in the OP about the aspects of AI I was referring to.

Very interesting and useful replies so far, everyone (with graphs!)—thank you.