#ai

See tagged statuses in the local Gatti Ninja community

putting wrong medical advice in their Overviews.

This is not funny. It is not acceptable. Today I saw someone who had asked a health-related question to Google Search. The AI Overview presented mixed-up, wrong information. I urged the person to ignore the AI Overview and use the regular site links.

They didn't know there were any regular site links, because the AI Overview filled essentially their entire window and they didn't know to scroll down. This is a common situation with busy, nontechie users.

They have depended on Google to point them at accurate information for so many years, and now Google Search spews out convincing looking AI garbage. This is not an anomaly.

Google's AI Overviews are full of wrong, partially wrong (even more dangerous!), and just plain misinformation. Answers that are completely reversed from supposed source pages because the AI …

In 2010, downloaded 70 GB of articles from JSTOR. He faced a $1 million fine and 35 years in prison
➡️ Aaron took his own life in 2013.

illegally downloaded over 80 terabytes of books from LibGen, Anna's Archive, and Z-library to train its models
➡️ facing no consequences.

Information is power, knowledge is power.

Beware and fight against those who want to keep you ignorant, unaware, and misinformed.

A black and white line drawing of Aaron Swartz with wavy hair, wearing a t-shirt and smiling.

Just read an article from an influencer musing on the Innovative power of AI.

Tasked with breaking the Enigma code, an AI system trained to recognise German using Grimm’s fairytales, utilizing 2,000 virtual servers, cracked a coded message in 13 minutes.

Let's pause for a second to let it sink.

And let's think for a second

Alan Turing “Bombes” could decipher two messages every minute.

😱. Suddenly the AI result isn't all that impressive any more.
AI cuts out all the research, knowledge gain, and insight. With all the resources available today, it still performs worse than a solution from 70 years ago (to be precise 26 times).
And this is seen as an impressive innovation 🤡🤯

"Sources":
Influencer post https://mastodon.social/@Caramba1/114470245795906227
Guardian article
https://www.theguardian.com/science/2025/may/07/todays-ai-can-crack-second-world-war-enigma-code-in-short-order-experts-say

Oh gosh, it's true, you really can enter a completely nonsense phrase into Google, ask for its meaning, and lo and behold, Google's AI will make shit up.

So if you've ever wondered what "to grow an avocado, you have to slap the squirrel" means, now you know.

Screenshot from a google search for "to grow an avocado you have to slap the squirrel meaning". "AI overview: The saying "to grow an avocado, you have to slap the squirrel" is a play on words, suggesting that the old method of dispersing avocado seeds, which relied on large mammals, is no longer effective due to their extinction. This is because avocados are an evolutionary anachronism, still requiring dispersal methods from before the decline of large mammals. Here's why: The Old Method: In the past, large mammals would eat avocado fruits, digest the seeds, and then disperse them in their droppings, effectively planting them in new locations. The Problem: Today, without large mammals to carry the seeds, avocados often rot where they fall or are eaten by smaller animals like squirrels. Humans as the Solution: Humans now take the role of dispersing avocado seeds through planting and cultivation. The saying highlights the evolutionary disconnect between avocados and the current ecosystem, emphasizing that their reliance on large mammals for dispersal is no longer sustainable." And then right at the bottom it says "Generative AI is experimental." No shit, Sherlock.

LINKioni - uno show intelligente

In questa puntata io e Joda, l’hacker di @storiespettinate, ci siamo riscoperti: emozionati per la prima puntata dello in un leggero.

Anche la nostra in cariola, costruita con chiodi e martello era emozionata.

Ospite Cristiano Dalianera, critico cinematografico e .

Ottima scusa per chiacchierare su questa arte da punti di vista sui generis (e anche sui suoceris)

https://castopod.it/@LINKioni/episodes/linkioni-17-aprile-2024

ci riprova e di nuovo vuole i tuoi dati per addestrare l’onnipresente !

Sta volta ha reso le cose più semplici per opporti, ma dando pochissimo tempo alla utenza: ha cominciato ad avvisare solo da poco. A maggio prende tutto quello che hai pubblicato e lo usa per addestrare il suo modello generativo.

Che poi mi ero già opposto nel 2024, ma niente, questi so de coccio.

Se vuoi opporti (di nuovo), ho aggiornato il semplice su come fare: https://artematiko.wordpress.com/2024/06/10/diritto-di-opposizione-a-meta-tutorial/

The German National Library @DNB_Aktuelles AI team just published a paper "Automatic Subject Cataloguing at the German National Library" in LIBER Quarterly (open access):

https://doi.org/10.53377/lq.19422

The paper documents in detail how they use open source tools such as Git, Data Version Control and Annif to implement automatic assignment of DDC classes and GND subject terms to digital publications in various formats. Great work!

Last night I Googled the Q. how many episodes of Would I lie to You has Victoria Coren Mitchell been on, having enjoyed a back episode where her & David Mitchell were on the team.

The Google/AI response was she was on every episode in series 15 & 16... which, as I had the series up on IPlayer was pretty easily demonstrated to be untrue.

While this is a pretty facile example, if it had been a more serious enquiry this level of inaccuracy could prove highly damaging!

AI is a problem!

A quando le che producono partite di calcio, gare, tornei sportivi artificiali?
(Con un fork per per le stesse, reali, ma che vengono fatte finire dall'ai come desidera lo spettatore)

(E poi: film che finiscono in modi alternativi, colloqui di lavoro idem, ecc)