You remember that dress? You know the one. Wanna feel old? That was 10 years ago. Anyway, if you’re nostalgic for those heady days of 2015 when people looked at something and were 100 percent sure they were right and the people who saw it differently were wrong, just ask someone their opinion on AI.
The Macalope has noticed lately that AI is a real divider. Many people now simply view it as normal and use it all the time. Certainly, lots of developers do but also, honestly, just everyday users. A lot of other people, however, see it as an utterly useless abomination, a harbinger of the downfall of society.
Today we’re going to figure out who’s right.
Okay, not really. Because the Macalope would like to posit that, much like the dress, they’re both right. BOOOOOOOOO!
Hey! You can’t boo the Macalope in his own column! Can you? How did that even get in here? And why can’t the Macalope delete it?
Look, AI has some great applications and can be a useful tool. It’s also sadly a business largely run by some of the worst people possible. Much like most major league sports, the game is not so much the problem as the business that’s running it.
Tech exec Anil Dash recently noticed how AI seems to drive people to fits.
It’s simply true that the Big AI platforms like ChatGPT are:
- Extremely bad for society, in many ways.
- Very genuinely popular with lots of people.
Let’s try to unpack this a little.
If you’d like to receive regular news and updates to your inbox, sign up for our newsletters, including The Macalope and Apple Breakfast, David Price’s weekly, bite-sized roundup of all the latest Apple news and rumors.
Foundry
AI detractors need to recognize that, despite the concerns, AI tools are providing value for many people. (Here, the Macalope is really just talking about LLMs.)
- “Vibe coding” may be an oversimplified marketing term, but AI coding tools make development easier.
- AI actually has an accessibility aspect to it that a lot of detractors don’t really want to talk about because social benefits don’t factor into their notion of AI as the knowledge equivalent of Juicero.
- Like it or not, AI results are faster and more convenient for average users who don’t have the time or capacity to digest large volumes of information.
Meanwhile, AI fans need to understand that there are some very good reasons for people to be distrustful and resentful of AI.
- AI is already being used (often just as an excuse) to lay off actual human beings, often leading to dubious results at best.
- AI companies routinely think nothing of scraping up whatever they want without concern for copyright or the labor of others in order to train their models.
- While sometimes exaggerated, there are real environmental issues to shoving rainforests into stoves in order to power server farms to churn out wrong answers to the simplest of questions.
- If we are, as many suspect, in an AI bubble, you can bet your bottom dollar that when it pops the wealthy investors in AI will head right to the Trump administration and demand your bottom dollar as part of a bailout of their over-investment of a technology that, while useful, never should have been jammed into literally everything.
You can obviously kind of see which side of this argument the Macalope leans toward.
Dash is not wrong that AI tools are being used by many people every day, but there are signs that it is not nearly popular enough to support the level of investment being poured into it, as The Economist notes (subscription required).
The Economist’s basic point is that no matter how much they might find AI tools convenient and useful, consumers are unlikely to shell out nearly enough for them in order for AI companies to make a return on the Dyson sphere-sized investment being made in them.
Businesses must do the rest.
And many signs indicate AI use in businesses is currently falling rather than rising.
Evidence is mounting that the current generation of models is not able to transform the productivity of most firms.
This brings us to Apple, the company still trying to figure out this whole AI thing. On Monday, surprising literally no one, Apple announced that its VP in charge of AI development, John Giannandrea, is retiring. Giannandrea will be replaced by Amar Subramanya, who was head of engineering for Google’s Gemini Assistant prior to leaving the company for Microsoft. It remains to be seen whether or not Subramanya can turn around the giant iceberg that is Apple’s AI effort to date.
The Macalope knows what he wrote.
Apple’s business practices are not always the best, particularly of late. And that’s too bad. Because the company has often been better than its peers on the environment, social issues, and privacy. And what’s needed most right now is a company that’s able and willing to do AI the right way. Until that happens, don’t expect opinions about AI to change.
