Monday Miscellany, 6/10/24

Will AI prove to be another impressive high-tech innovation that fizzles, or at least has far less impact than expected?  Probably not.  As my techie friend keeps telling me, it will get better.  But right now, the massive investments in AI are facing problems.

Christopher Mims tells the story in his Wall Street Journal article The AI Revolution Is Already Losing Steam with the deck “The pace of innovation in AI is slowing, its usefulness is limited, and the cost of running it remains exorbitant.”

The article is behind a paywall, but here are some of the problems, from a business point of view, that Mims describes.  In general, he says, despite all the money being poured into the technology, improvements are slowing.  Plugging in ever-more data has resulted in better performance, but there is a limit to the amount of data that is available.  Engineers have tried to make progress by using AI to generate “synthetic data” for AI, but that isn’t working.

Also, contrary to expectations, AI hasn’t resulted in increases in productivity, much less reduced payrolls.  This is because companies need to hire someone to check the AI-generated information, due to the phenomenon of “hallucinations” (wrong answers that AI keeps presenting as factual, which as I’ve blogged about happens virtually every time I’ve tried to use the technology).  To get better results, companies need employees with the special skill of constructing AI prompts, which adds someone else to the payroll.  While many companies are now using AI, they are mainly using it to help workers, not replace them.

And the technology is super expensive.  Nvidia, which makes chips for AI, is certainly doing well, showing the huge amount of investment being made in the technology.  But the returns are much less.  Last year, AI developers, says Mims, spent $50 billion on chips, but brought in only $3 billion in revenue.

Furthermore, says Mims, “the costs of running it far exceed the already eye-watering cost of training it. That’s because AI has to think anew every single time something is asked of it, and the resources that AI uses when it generates an answer are far larger than what it takes to, say, return a conventional search result.”

Concludes Mims,

These factors raise questions about whether AI could become commoditized, about its potential to produce revenue and especially profits, and whether a new economy is actually being born. They also suggest that spending on AI is probably getting ahead of itself in a way we last saw during the fiber-optic boom of the late 1990s—a boom that led to some of the biggest crashes of the first dot-com bubble.

The Internet Comes for Amazon Tribes

The Marubo people live deep in the Amazon rainforest.  It takes a week to reach their villages of huts, and their way of life has not changed for hundreds of years.  But now they have high-speed internet.

Brazil has made Elon Musk’s satellite-based internet system Starlink available to hundreds of remote jungle tribes.

A story in the New York Times, The Internet’s Final Frontier: Remote Amazon Tribes, tells the tale.  It chronicles how after only nine months, the Marubo, after resisting the modern world for generations, are now struggling with teenagers hooked on screens, social media gossip, violent video games, misinformation, scams, and pornography.  For all of its benefits, the technology that has ravaged advanced Western cultures is now ravaging theirs.

The story includes a priceless quotation that sums up how just about everyone in the world feels about the internet:

“When it arrived, everyone was happy,” said Tsainama Marubo, 73, sitting on the dirt floor of her village’s maloca, a 50-foot-tall hut where the Marubo sleep, cook and eat together. The internet brought clear benefits, like video chats with faraway loved ones and calls for help in emergencies. “But now, things have gotten worse,” she said.

She was kneading jenipapo berries to make a black body paint and wearing ropes of jewelry made from snail shells. Lately, the youth had become less interested in making such dyes and jewelry, she said. “Young people have gotten lazy because of the internet,” she said. “They’re learning the ways of the white people.”

Then she paused and added, “But please don’t take our internet away.”

Pro-Life Protesters Get Long Prison Terms

The Biden Administration is throwing the book at Pro-Life protesters.

A 30-year-old woman, Lauren Handy, was sentenced to 57 months in prison–nearly five years–plus three years of supervised release after being convicted of violating the Freedom of Access to Clinic Entrances Act and conspiracy against rights.  She and other protesters had blocked the entrance to an abortion clinic in Washington, D.C., back in 2020.

Others at the protest were also given long prison terms:  Jonathan Darnel (34 months), Herb Geraghty (27 months), Jean Marshall (24 months), and Joan Bell (27 months).

See How Much Jail Time?  Here Is a List of Prison Sentences to see how these sentences for trying to save the lives of babies line up with other crimes.  Manslaughter, for example, will get you a sentence of 5 years.  So will robbery with a deadly weapon.  Kidnapping will send you in the slammer for 36 months.