Welcome to our annual checklist of the worst, least profitable, and easily dumbest applied sciences of the yr.
We wish to suppose there’s a lesson in each technological misadventure. However when know-how turns into depending on energy, generally the takeaway is easier: it might have been higher to remain away.
Regrets—2025 had just a few. Listed below are a few of the extra notable ones.
—Antonio Regalado
A short historical past of Sam Altman’s hype
Every time you’ve heard a borderline outlandish concept of what AI will probably be able to, it usually seems that Sam Altman was, if not the primary to articulate it, at the very least probably the most persuasive and influential voice behind it.
For greater than a decade he has been recognized in Silicon Valley as a world-class fundraiser and persuader. All through, Altman’s phrases have set the agenda. What he says about AI is never provable when he says it, but it surely persuades us of 1 factor: This highway we’re on with AI can go someplace both nice or terrifying, and OpenAI will want epic sums to steer it towards the proper vacation spot. On this sense, he’s the final word hype man.
To grasp how his voice has formed our understanding of what AI can do, we learn nearly the whole lot he’s ever mentioned concerning the know-how. His personal phrases hint how we arrived right here. Learn the total story.
—James O’Donnell
This story is a part of our new Hype Correction package deal, a set of tales designed that will help you reset your expectations about what AI makes doable—and what it doesn’t. Take a look at the remainder of the package deal right here.
Can AI actually assist us uncover new supplies?
One among my favourite tales within the Hype Correction package deal comes from my colleague David Rotman, who took a tough take a look at AI for supplies analysis. AI may rework the method of discovering new supplies—innovation that may very well be particularly helpful on the planet of local weather tech, which wants new batteries, semiconductors, magnets, and extra.
However the discipline nonetheless must show it could actually make supplies which might be really novel and helpful. Can AI actually supercharge supplies analysis? And what would that appear like? Learn the total story.
