Something That Gives Me Hope with AI

Right now, if you are an AI cynic, it’s pretty easy to feel hopeless. Several AI companies are announcing massive IPOs, thousands of layoffs have been blamed on AI and it seems like there are no brakes on the AI train.

The legal front has also been largely devoid of good news. Sure, Anthropic agreed to a $1.5 billion settlement with authors, but that only dealt with the piracy of ebooks, not the actual training. In that case, the court ruled the actual training was legal and protected under fair use.

While that’s just the first ruling in what will likely be many, it doesn’t bode well for human creators.

Then you have Google pushing more and more people to AI searches, which have been a major blow to the open web. Sites like this one, who have already seen massive drops in traffic are likely to lose even more.

And that says nothing about the bots on social media, the rise of AI-generated content on YouTube and just the general prevalence of AI on the internet.

So what is an AI cynic to do? Well, there is one thing that they can take hope in, that they are not alone and, regardless of their demographics, they are like in the majority.

The Backlash Against AI

A March 2026 article by Michelle Faverio and Emma Kikuchi at Pew Research Center compiled a series of very interesting surveys and studies about how the public feels about AI.

In a survey of Americans, roughly half said that they were more concerned than excited about AI. Another 38% said they were equally excited and concerned.

Those numbers have remained relatively consistent over the past three years. Meaning that, even after three years of AI hype and technological advances, the public is still extremely wary.

In a similar survey, Pew asked about how AI will impact different areas, ranging from the economy to the criminal justice system. Only one area, medical care, had a plurality of people who felt the impact would be positive. In other areas, it was decidedly negative. For example, in personal relationships, only 7 percent said it would be positive while 47% said it would be negative.

Though more people are using AI in the workplace, that number is still just at 21%. The percent of people using AI at work either “Not Much” or “None” also went up, reaching 65%. What dropped was the group who had not heard of AI use. 

That is to be expected as awareness of AI has continued to grow. As of last year, just 5 percent of those surveyed said they had not heard of AI.

Despite that growing awareness, US adults are still much more wary of AI than AI experts. AI experts were almost four times more likely to say that AI will have a positive impact than a negative one. Meanwhile, the general public was twice as likely to have a negative view than a positive one.

In short, though awareness of AI has grown, excitement for it has not. However, you didn’t need surveys to know that; you just had to look at the news.

The Business Pullback

There’s no better microcosm of this AI backlash than Microsoft. Microsoft has not only invested heavily in its own AI products, but also in OpenAI. As part of that, they have pushed their Copilot AI assistant into just about everything they can.

This earned Microsoft the nickname “Microslop”, which the company tried (and failed) to distance itself from. However, even with the leverage of Windows and Microsoft 365, fewer than 3.3% of their paying commercial seats are paying for Copilot. According to some metrics, Microsoft has actually lost ground in the AI race, dropping from 1.5 percent to just 1.1.

But even the customers who do pay likely cost Microsoft money. Simply put, the economics of AI, at least right now, are horrible. As this article by Victor Tangermann at Futurism points out, free and low-cost AI tools may be dying out in favor of pay-as-you-go models. The free and flat-rate plans that have survived are getting new caps and price increases.

Because of this, companies that once pushed employees to use AI as much as possible are now setting limits. Even the most full-throated AI supporters are now pulling back on the reins. The halcyon days of 2024 are over. The economics of AI were always dubious, but now the bills are starting to come due.

As bad as all of this is for AI, they aren’t likely to fare much better with the next generation of workers. For all the reputation students have for “ChatGPTing their way through school,” that’s not really the case.

You’re All Wrong and The Kids Are Right

If you’ve been following news around graduation season, you’ve likely heard about the multiple graduation ceremonies where students booed and heckled commencement speakers mentioning AI. 

This included former Google CEO Eric Schmidt and Big Machine Records founder and CEO Scott Borchetta. Meanwhile, Delta’s CEO, Ed Bastain, received cheers and applause when he mentioned that he set aside AI and “put pencil to paper” for his speech.

Many were surprised by this, but if you look at the data, it’s much less surprising. Going back to the Pew article, it found that just under half of 13-17-year-olds used AI to help with schoolwork. Another 23 percent said they only used it a little and another 21 percent said they used it some. Only 10 percent said they did it all or most.

While that sounds bad, compare it to the numbers from before AI. In 2010, a study found that 40 percent of students admitted to having copied at least some content without attribution. In short, AI didn’t create many new cheaters; instead, it just gives those who might have done it anyway an easier way to do it.

While self-reported surveys are always to be taken with a grain of salt, the apples-to-apples comparison seems to indicate that, while AI cheating is absolutely a problem, it’s not much different in scope than copying from the internet was 16 years ago.

Is AI cheating more complex? Of course. Are there serious challenges for educators? Absolutely. But it shouldn’t be a shock that many students aren’t cheating and that many of those who are only did it reluctantly. Even among the people who use AI, AI is not very popular. 

That is true both in and out of the classroom.

Bottom Line

A February article by Julius Nyerere Nyambok highlighted something I, and many others, have lost sight of. Simply put, 84% of humans have never used AI. While the exact number varies from country to country, only in a few places, such as the UAE and Singapore, do AI users outnumber non-users.

But even those using AI aren’t making significant use of it. Though 16% of the world has used a free AI chatbot, only 0.3% have paid anything for it. AI feels ubiquitous; it feels like it’s everywhere and that it is inevitable. However, it’s not. 

AI remains extremely unpopular. Increases in awareness, usage, and in the technology have not translated to greater acceptance. That gives me at least a degree of hope.

None of this is to say that AI is doomed or that it will collapse. What it is saying is that the future hasn’t been written yet. Though extremely powerful forces are pushing AI, they haven’t been able to move the needle when it comes to public opinion.

At some point, AI has to answer to the public. That opinion may not matter much right now, but if AI is going to truly be a game-changing long-term technology, it needs the public on its side.

That is not happening. At least not yet.