Inevitably, these conversations will take a turn: AI is having all these ramifications nowbut as technology advances, what happens next? Usually when they look at me and expect a prediction of either doom or hope.
I'm probably disappointed, if only because AI is becoming increasingly difficult to predict.
Nevertheless, MIT Technology Review I have to say that we have a pretty good track record of understanding where AI is going. We just published a list of sharp predictions for what's next in 2026 (you can read my thoughts on legal battles over AI). And all of the predictions on last year's list came true. But with each holiday season, figuring out the impact of AI becomes increasingly difficult. The main reason for this is three big unanswered questions.
First, we don't know if large-scale language models will continue to get progressively smarter in the near future. This particular technology underpins almost all of the excitement and anxiety around AI today, powering everything from AI companions to customer service agents, so its slowdown would be a pretty big deal. In fact, it's such a big deal that we spent an entire series of articles in December talking about what the new post-AI hype era will look like.
Second, AI is quite unpopular among the general public. This is just one example. Almost a year ago, OpenAI's Sam Altman stood next to President Trump and excitedly announced a $500 billion project to build data centers across the United States to train ever-larger AI models. They neither guessed nor cared that many Americans would be adamantly opposed to building such data centers in their communities. A year later, Big Tech is fighting an uphill battle to win public opinion and keep building. Can we win?
Lawmakers' reactions to these complaints have been deeply confusing. President Trump has delighted Big Tech CEOs with his move to make AI regulation a federal rather than state issue, and tech companies now want to make it law. But the crowd that wants to protect children from chatbots, ranging from California progressives to the increasingly Trump-aligned Federal Trade Commission, each has distinct motives and approaches. Will they be able to put aside their differences and rein in AI companies?
When a depressing holiday dinner table conversation gets this far, someone will probably say something like, “Hey, isn't AI being used for objectively good things?” Making people healthier, unearthing scientific discoveries, and better understanding climate change?
Well, in a way. Machine learning, an older form of AI, has long been used in all kinds of scientific research. One branch, called deep learning, forms part of AlphaFold, a Nobel Prize-winning protein prediction tool that has transformed biology. Image recognition models are getting better at identifying cancer cells.
