π Newsletter #011 - Failing Grades, Mass Delusions, and Hiring Back Humans
Original images by CHUTTERSNAP and Larry Costales on Unsplash; glitched and edited.
β¨ In This Edition β¨
- π Short-Form Content: Failing grades, export bans, and workplace mass delusions - and yet another reason to ditch Chrome.
- π Good News!: Companies can be found liable for LLM-generated content; Pubs are banning AI imagery; Ford is hiring back humans.
- π― The Post-Script: Reading Cory Doctorow's new book, The Reverse Centaur's Guide to Life After AI.
π Short-Form Content
Chatbots ("AI")
- Students are raking in more failing grades as professors see greater AI usage & dwindling math skills. It's almost like outsourcing all our thinking is not the best idea?
- Microsoft had to shut down 70+ of its GitHub repos after hackers implemented malware in AI coding software, for example, Claude.
- Here's a cool infographic on Who's Suing Whom in AI.
- KPMG wrote a report on how AI has transformed industry. Only problem, they used AI to write the report, and it made up a lot of the evidence. The report has now been withdrawn.
- New Anthropic models have been hit with export bans. Switching your company to rely on one of those huge, corporate AI systems for everything, baking it into the fibres of every PR under the sun, means that you are susceptible to someone turning off the tap. File under: "Digital Sovereignty."
- This excellent essay on "our workplace LLM mass delusion," by ava - here's a small excerpt:
How all the gates and wallets have been opened for this utter disappointment, showing that the obstacles for implementing anything thought to be inherent and unavoidable in the organization are just a fluke, a lie, an arbitrary thing? How it all created a culture of feeling repeatedly gaslit over months about this whole assessment, as if you must be the one that is insane?
- Websites are being scraped at alarming, unsustainable rates by AI to ingest literally everything - and then sell it back to us, overwhelming maintainers. No no, this will be fantastic for the future of the web and FOSS projects.....
- "If AI could think creatively, then it should be able to train itself on its own output and would have no need to steal the collective creative output of humanity. And if creative output has no intrinsic value, then they have built themselves a machine worth nothing." by Lana.
- McSweeney's is once again hitting hard - AI Economics for Dummies:
Acquiring one grape costs Alex $2 billion. Alex offers to sell Mike one grape a month for the next 12 months for $1 billion per grape. Alex asks for the full $12 billion up front and provides Mike with one grape for the first month. Alex makes a $10 billion profit this month; his ARR is $120 billion, and his profits are trending up at an infinite rate. The Wall Street Journalβs business editor moves into Alexβs house, having accepted a part-time position as Alexβs human footstool. He never asks to see the books.
- It is shockingly easy to manipulate AI search, for example via just a couple targeted reddit posts. I'll leave it as an exercise for the reader to figure out why this may be a problem in the contexts of democracy, elections, and the political sphere on the whole.
- Here's a neat survey tool to see where you actually stand on AI: The AI Compass. (I got "The Skeptic!")
- Guy Who Sucks at Being a Person Sees Huge Potential in AI.
Non-Chatbots
- Also file under "Digital Sovereignty:" Facebook is paying non-Albertans to promote Alberta separatism. Accounts are apparently run from places like Indonesia, Pakistan, India, and others. Isn't that interesting, for such a ""home-grown"" initiative as this one!?
- Hey, remember Bill C-22? Debate on it was weirdly cut very short. Despite many organizations opposing it and pushback on the whole being significant.
- It is still so weird to me that governments the world over are attempting to regulate and ID every person under the sun, rather than holding the platforms accountable and demand that they improve.
- Hey, so remember Manifest v2 for Chrome? And how that meant that popular ad-blockers would no longer work, and you probably had to install a different one - one specifically tailored to Manifest v2? So yeah, they're disabling all of that now.
- If you're still using Chrome, my recommendation is to remove it from your devices, like, today (back up your bookmarks etc if you need to) and switch to one of the very capable alternatives.
- Dynamic pricing is just gross.
π Good News!
- A court in Germany has ruled that Google is liable for the content its AI overview spits out.
- This means that AI overviews are being brought somewhat closer to the level of scrutiny of, well, basically any other service that isn't AI.
- Google's defense, which sounded a lot like "well, everyone knows that the AI overview is bs," somehow didn't really do the trick.
- Will this change their plans of effectively killing off Search and replacing it with generative "content?" I guess we'll see.
- While a lot of very rich people with a lot of skin in the game are trying to tell us that there is no alternative and everyone's totally on board with AI actually, Steve Wozniak got a lot of cheers after telling grads that they "have AI - actual intelligence."
- Game dev engine Godot updated their Contribution Policies such that vibe-coded slop is no longer welcome.
- There's this cool pub in Dublin, Thomas House, that is banning all AI-generated posters and flyers from their pub. You love to see it!
- Ford is hiring back humans after they
~~pivoted~~hard to AI - shockingly enough, it turned out that the AI was no match for human skill.
π― The Post-Script
- Reading Cory Doctorow's new book, The Reverse Centaur's Guide to Life After AI.
Follow this blog via RSS | Find me on Mastodon | Bubbles
Recent Posts