Wednesday, March 22, 2023

The AI hype bubble is the new crypto hype bubble

Each of us carries around little pockets of guilt, misdemeanors camouflaged and unsuspected even by our dearest. They remain dormant until roused, jab our conscience and are forgotten until stirred again.


Rozenshtein, Alan Z., Moderating the Fediverse: Content Moderation on Distributed Social Media (November 23, 2022). 2 Journal of Free Speech Law (2023, Forthcoming), Available at SSRN: https://ssrn.com/abstract=4213674 or http://dx.doi.org/10.2139/ssrn.4213674


No other council has bad apples like this …

The AI hype bubble is the new crypto hype bubble



Pluralistic / Cory Doctorow: “…Blockchain was a solution in search of a problem. So is AI. Yes, Buzzfeed will be able to reduce its wage-bill by automating its personality quiz vertical, and Spotify’s “AI DJ” will produce slightly less terrible playlists (at least, to the extent that Spotify doesn’t put its thumb on the scales by inserting tracks into the playlists whose only fitness factor is that someone paid to boost them). 

But even if you add all of this up, double it, square it, and add a billion dollar confidence interval, it still doesn’t add up to what Bank Of America analysts called “a defining moment — like the internet in the ’90s.” For one thing, the most exciting part of the “internet in the ’90s” was that it had incredibly low barriers to entry and wasn’t dominated by large companies – indeed, it had them running scared. The AI bubble, by contrast, is being inflated by massive incumbents, whose excitement boils down to “This will let the biggest companies get much, much bigger and the rest of you can go fuck yourselves.” Some revolution. AI has all the hallmarks of a classic pump-and-dump, starting with terminology. AI isn’t “artificial” and it’s not “intelligent.” “Machine learning” doesn’t learn. On this week’s Trashfuture podcast, they made an excellent (and profane and hilarious) case that ChatGPT is best understood as a sophisticated form of autocomplete – not our new robot overlord…”




One Useful Thing – Power and Weirdness: How to Use Bing AI – Bing AI is a huge leap over ChatGPT, but you have to learn its quirks. Most people know about Bing AI from the chatbot’s well-reported week of doing weird and threatening things. I think that those stories focused people on the less-useful (though admittedly more scary) parts of the Bing experience.

 Now, Bing has since been improved considerably, and (good news, everyone!) it no longer wants to split up your marriage. Instead, it is the most powerful AI tool out there, and can far outperform ChatGPT in some areas. But it has some definite quirks to be aware of, so I wanted to offer some tips on how to use it to do some pretty amazing things, while avoiding some of the traps. Your starting model for using Bing should be ChatGPT, as both AIs started with OpenAI’s GPT models. So the rules you have been learning for ChatGPT apply to Bing as well, so, as I discuss in my guide on how to use ChatGPT to improve writing, you need to carefully consider the prompts you are creating, and work interactively with Bing to get the best results. But Bing adds something special – it is connected to the internet. So the first rule of using Bing is…Always ask it to look things up..”


 Twitter v. Mastodon v. Post v. Other Possibilities by Teri Kanefield

1. “Twitter – Twitter has done a lot of good. It allowed communities to form. It allowed marginalized voices to be heard. It allowed crucial information to be disseminated. It even saved lives. Like Facebook and other platforms that rely on algorithms to stimulate engagement, Twitter has also done a lot of harm.

The Pew Research Center says this: Nearly all the content people see on social media is chosen not by human editors but rather by computer programs using massive quantities of data about each user to deliver content that he or she might find relevant or engaging. This has led to widespread concerns that these sites are promoting content that is attention-grabbing but ultimately harmful to users – such as misinformation, sensationalism or “hate clicks.”

Musk’s only “power” is the power we give him. He is not the president of the United States. He isn’t a world leader or elected official. He controls the Twitter code. That’s his power. Some Twitter users believe that Musk will crash and Twitter will outlive him and something different will replace the Twitter of today. This seems unlikely to me, but I suppose it’s possible. I also suspect a lot of people are keeping their Twitter accounts active while simultaneously building a home elsewhere as they watch what develops. (I’ve heard that it’s best not to delete a Twitter account, particularly if you are verified: Someone else can reactivate your account within 30 days or use your name. If you want to leave entirely, the best option is to make your account inactive while retaining control.)…”