It took less than a week for the newly available ChatGPT technology to get a million monthly active users (MAU) in November of 2022, making it the most quickly adopted technology at that time. Now, according to analyst firm Sensor Tower, OpenAI has hit the one billion mark, even though if you listen to people on the street, you might hear a lot of negative things about AI.
In terms of sentiment, some point to recent commencements at American universities, where attendees moaned and groaned about the AI revolution, and surveys where people say they deliberately don’t use AI at work, for ethical reasons. You might also hear a lot about rug pulls, job displacement, and, thanks to the very public kerfuffle between Anthropic and DoD, fully autonomous lethal weapons.
So why is GPT now getting McDonalds numbers?
Flying Business
One reason for the enduring popularity of LLMs might be business applications. In 2025, OpenAI reported that its lineup had hit a million business customers, with over 800 million personal users. That’s weekly, not monthly.
“We’re proud to work with category leaders in industries like financial services, healthcare, retail, and more, where our technology is making intelligence central to their customer experiences, internal operations, and team-level workflows,” spokespersons wrote, at the time, around the three-year mark, as they celebrated this milestone.
That’s one reason why AI continues to snowball even as many people want to fight its incursions into their lives. But it’s just one aspect of a disconnect that you might say comes from the old saying “do as I say, not as I do.”
The Limited Power of Sentiment
One of the basic ideas here is that a powerful and universal trend compels people to get involved, because at the end of the day, we are all in this life together. Someone might have misgivings about something, but go along with their team, or their family, or their friends.
I was reading this report from CNBC written June 12, the same day I was reading it, where writer Matthew Chin chronicled a lot of recent happenings, contrasted sentiment to practice, and repeatedly argued that pushback isn’t going to successfully culminate in any substantial arrest of AI development. Let me just pick out a couple of lines:
“Yet despite mounting public backlash, global AI usage has surged to record highs.”
This is just reporting what happened. But then you have this:
“While ethical user considerations may have driven usage in cases like Anthropic’s feud with the Defense Department, analysts say the broader trajectory of AI usage is less likely to be derailed by sentiment.”
Chin does name and quote an analyst as ballast:
“’The strong trajectory of AI adoption shows no sign of slowing,’ said Hanno Stegmann, managing director and partner at the Boston Consulting Group’s AI and technology team, BCG X.”
That’s one such assertion. Then, later, Chin writes:
“As AI plays an increasingly central role across daily life, any souring public sentiment will likely make few dents in overall usership.”
Chin then cites a study, and quotes Abe Yousef from Sensor Tower, that same firm that enumerated GPT’s landmark, as saying:
“’While negative sentiment towards AI… is undeniably growing, consumers are increasingly using and relying on these platforms.’”
Why am I including all of this verbatim? Because I think it gives you the flavor of how this hits, as a reader, at least to me. It’s credible. It’s compelling. I find myself wondering: can people really choose their own adventures in this area of their lives, especially at work?
Some Tips
For what it’s worth, individuals with concerns can get guidance on how not to over-rely on AI. I liked this piece by our own Forbes contributor Sol Rashidi, which includes bullet point tips like “stay curious,” “limit dependency,” and my personal favorite, which we should all do anyway: “practice reflection.”
I sit in a lot of conferences on AI, in Boston, at places like Davos, Switzerland, and around the world. I represent part of the MIT community, and I hear lots of voices. I don’t think there’s one monolithic finding, or one precise path determined, for AI. But the question of sentiment is a big one. There’s a growing consensus that in the age of agentic AI, and the nearness of AGI or super-intelligence, we need to figure out more about what it means to be human. Because apparently, all of us will be using LLMs, or whatever comes next.
