(C) Daily Kos
This story was originally published by Daily Kos and is unaltered.
. . . . . . . . . .



AI Hype is the Dot.com Bubble, not the Crypto Bubble [1]

['This Content Is Not Subject To Review Daily Kos Staff Prior To Publication.']

Date: 2023-06-12

This is pretty interesting, if a bit overhyped in the article:

Sorting is so basic that algorithms are built into most standard libraries for programming languages. And, in the case of the C++ library used with the LLVM compiler, the code hasn't been touched in over a decade. But Google's DeepMind AI group has now developed a reinforcement learning tool that can develop extremely optimized algorithms without first being trained on human code examples. The trick was to set it up to treat programming as a game. AI system devises first optimizations to sorting code in over a decade | Ars Technica

You can see a lot of caveats in the paper itself and in various discussions about the paper online. Simply, the changes may be only really applicable to a specific architecture and they diminish significantly as the numbers in the sort rise. Having said that, this is still an interesting look at how real-world problems can be solved by machine learning algorithms. And that highlights both the peril and the promise of AI and what makes it different than the last two tech hype cycles.

In the last two years, we have been subjected to an enormous amount of hype around the so-called metaverse and the use of blockchains in everything up to and including seemingly breakfast cereal. In both cases, they were all hype. At no point did either ever offer a realistic path to useful products.

Versions of the metaverse have been around since the early 2000s if not earlier -- Second Life, anyone? -- and they have all been relegated to niche products at best. Simply put, VR is bad at creating truly immersive, interactive experiences when those experiences must include other people because they cannot mimic true physical interactions and likely never will. And since they cannot do that well, they become inefficient experiences that requires you to wear uncomfortable headgear for long periods of time. A VR meeting, after all, is really not that much more immersive than a zoom meeting with the added disincentive of wearing a stupid headset for an hour. And that doesn't even get into motion sickness or vertigo issues that a significant portion of people who use VR experience. Outside of a certain style of games, there is almost always a better, more efficient, more comfortable means of achieving the desired outcome.

Blockchain or crypto is even worse. That never had a real-world use case outside of speculation and maybe blackmail. Blockchains are slow, they have no means of transaction rollback, making them useless for the vast majority of real-world situations, they are easily subject to abuse and fraud, and crypto currencies enforced scarcity makes them useless as a means of commerce. There was literally no reason for them to exist at all outside of hype and maybe the desire to do crimes and not get caught (and even that is an overblown aspect of crypto currencies.)

AI, or what we call AI since none of this stuff is any more intelligent than your average paving stone, has real applications. It can do real work and it can really improve how certain task are performed. That has promise and danger. It can make lives better or ruin them. A lot of how this plays out may depend upon how we as a society deal with the externalities of AI and determine who controls what it is and is not allowed to do and influence.

In that way, it is very much like the fist dot.com boom. The dot.com era was ridiculous. It gave us dumb websites and overhyped brands and morons trying to get rich quick setting billions of dollars alight and, eventually, a recession. But it also gave us new programming languages, untold miles of high-speed internet cables, and a new communication paradigm that we have used for good and ill ever since. It created real things amongst all the nonsense and hype.

This latest AI craze is almost certainly more hype than reality. Most of the things the AI boosters claim will happen almost certainly won't -- the tools are just not anywhere near as competent as they claim and likely won't be before the money runs out. But these algorithms will create some real, effective programs that do help perform real tasks. And that will have real effects on real people, for good and ill.

We owe it to ourselves to ignore the hype and actually pay attention to what these tools and companies are doing, where they are succeeding and failing, and who is paying the cost for both. We shouldn't just sit back and assume that the people who build them should be allowed to "innovate" on the backs of society, and we shouldn't allow them to sell us a fantasy of all powerful AI coming to steal our souls and crush our bodies beneath a robot boot. We should instead understand that buried in the self-serving hype are attempts to force society to allow them to exploit people for as much money as they can, consequences be damned, and the kernels of some things that can bring actual benefit to society as a whole. We should, that is, treat them like every other industry that has ever existed -- regulate them to control the damage they can do and ensure that the benefits are widely available and shared.

AI, unlike the metaverse and the blockchain, has some potential utility. That, unfortunately, means more work on our part as a society -- but is work we likely need to do. Otherwise, we will end up with another Facebook. Or worse, another Pets.com puppet.

[END]
---
[1] Url: https://www.dailykos.com/stories/2023/6/12/2174360/-AI-Hype-is-the-Dot-com-Bubble-not-the-Crypto-Bubble

Published and (C) by Daily Kos
Content appears here under this condition or license: Site content may be used for any purpose without permission unless otherwise specified.

via Magical.Fish Gopher News Feeds:
gopher://magical.fish/1/feeds/news/dailykos/