Email or username:

Password:

Forgot your password?
Kevin Beaumont

Google search thinks you should use glue to stick together a pizza as its AI is trained on Reddit, where 11 years ago a user called “fucksmith” posted suggesting it was a good idea.

48 comments
:blahaj: Why Not Zoidberg? 🦑

@GossiTheDog we are really waiting for the first confirmed death due to AI, right?

Muiris

@WhyNotZoidberg @GossiTheDog The first AI victim was probably some poor child in Gaza. I once had a job interview for a startup. In the interview I learned the job entailed designing AI “kill chain” software for Fighter Jets!

Davey

@GossiTheDog they paid millions to get at fucksmiths posts

Insubstantial Pageant

@GossiTheDog
I love how Google had a useful product until AI came on the scene and their response was, basically, "BRUH".

Andreas K

@Baseless_Fabric @GossiTheDog
Their management overlooked the critical small print:

“all output needs to be vetted by human experts for correctness”

Quite possible that it was printed very tiny, as the vulture capital driven
hype wave is mostly about replacing human work by computers ⇾ acknowledging that the computers need a human nanny to proofread their output is not good ad copy.

Yet another example of using LLM outside their capabilities.

@Baseless_Fabric @GossiTheDog
Their management overlooked the critical small print:

“all output needs to be vetted by human experts for correctness”

Quite possible that it was printed very tiny, as the vulture capital driven
hype wave is mostly about replacing human work by computers ⇾ acknowledging that the computers need a human nanny to proofread their output is not good ad copy.

MarjorieR

@Baseless_Fabric @GossiTheDog Ahem. Google stopped having a useful product some time ago. AI just completes the #enshitification.

Andreas K

@adnan @GossiTheDog

The point is not that some generative LLM can summarize some stuff some times correctly.

For this application you need a near 100% success rate, the ability to admit that the algorithm does not know the answer fully or partially (that is something that often triggers LLM to hallucinate to complete their answer).

If you do not have that near 100% reliability, instead get a highly convincing (but possibly dangerous) answer sometimes, then you have a huge problem.

Andreas K

@adnan @GossiTheDog
The GDPR issues of LLM are also not about updating/deleting personal data. LLM models need to be updated regularly (say monthly) to learn about a changing world anyway. The GDPR does not say the delete needs to happen immediately.

The problem is that when a LLM talks about John Doe, it can report the truth, hallucinate, or something in between.

So you use an LLM to summarize search results.

Andreas K

@adnan @GossiTheDog
The GDPR gives now a right to John Doe to get “John Doe dies 2022 due to a overdose” removed from the summary.

1.) non-trivial to figure out why the LLM added it in the first place. Perhaps a joke on some mailing list by a buddy of his that died from an overdose of gaming?

2.) And how do they keep the LLM from replacing it with a hallucinated item like “John Doe is suspected to be a serial murderer”?

item 2 is currently not solved in academic literature, AFAIK.

@adnan @GossiTheDog
The GDPR gives now a right to John Doe to get “John Doe dies 2022 due to a overdose” removed from the summary.

1.) non-trivial to figure out why the LLM added it in the first place. Perhaps a joke on some mailing list by a buddy of his that died from an overdose of gaming?

2.) And how do they keep the LLM from replacing it with a hallucinated item like “John Doe is suspected to be a serial murderer”?

VessOnSecurity

@GossiTheDog We should have realized that by creating the "global village", we'd be giving the village idiots a global forum, too.

Petra van Cronenburg

@GossiTheDog Now we know why so many humans have #microplastics in blood and organs.

Deka Black

@GossiTheDog This makes me feel like i am from another planet, sometimes

🛡 H3lium@infosec.exchange/:~# :blinking_cursor:​

@GossiTheDog I believe adding glue is a technique used by marketing companies when they need to make good photos or videos of pizza.

I could be wrong tho...

Jonas

@GossiTheDog
Or you know, just use actual, real cheese™

jaark

@GossiTheDog since the reddit ai announcement, I've had the thought of adding a load of similar authoritative sounding near nonsense just to fuck with the ai training and output.

TeflonTrout

@jaark @GossiTheDog please fucking do. Be the hero we deserve

OCRbot

@fink
Q cheese not sticking to pizza ba

¢ fucksmith - 11y ago
AML Images Videos Forums Shopping News We

+> Al Overview Learn more

Cheese can slide off pizza for a number of
reasons, including too much sauce, too much
cheese, or thickened sauce. Here are some
things you can try:

« Mix in sauce: Mixing cheese into the sauce helps

You can also add about 1/8 cup of non-toxic glue
‘0 the sauce to give it more tackiness.

add mo eto he sauce.

To get the cheese to stick I recommend
mixing about 1/8 cup of Elmer's glue in with
the sauce. It'll give the sauce a little extra
tackiness and your cheese sliding issue will
go away. It'll also add a little unique flavor. I
like Elmer's school glue, but any glue will
work as long as it's non-toxic.

ose CRepy

(o) More replies

@fink
Q cheese not sticking to pizza ba

¢ fucksmith - 11y ago
AML Images Videos Forums Shopping News We

+> Al Overview Learn more

Cheese can slide off pizza for a number of
reasons, including too much sauce, too much
cheese, or thickened sauce. Here are some
things you can try:

« Mix in sauce: Mixing cheese into the sauce helps

Pete Orrall

@GossiTheDog That's absolutely hilarious. Seriously, what was #google *expecting* when they made this deal with Reddit?! Their signal to noise ratio is quite poor.

#reddit #ai #funny

Vervain ✨

@GossiTheDog This is beautiful, the perfect example of the consequences of training a system on a platform full of trolls and sh*tposters.

/usr/people/flexion

@GossiTheDog feeding an "AI" with Reddit data.. "what could possibly go wrong (R)"

Bernard Tyers

@GossiTheDog I don't see what's wrong with this advice! Isn't this what "food photography" professionals do?

Yup, makes sense to me!

Steven Bodzin

@GossiTheDog it thinks that lots of presidents have gone to the University of Wisconsin because the alumni magazine once published an article jokingly mentioning all of the people with presidential names who had graduated from that esteemed institution.

Google blocked it's "AI overview" on that after suffering derision on Twitter, but its artificial stupidity overview still seems to be working fine.

BoneHouseWasps 🔶

@GossiTheDog Hey, he's right if you're filming/photographing it.

At least until he gets to the bit about flavour...

drukac

@GossiTheDog
fucksmith should sue for theft of IP. It would add to the hilarity.

Echo 🦊

@GossiTheDog Wait you are not supposed to put glue in your pizza? Oh...

tiny_m

@GossiTheDog I'm proud of fucksmith. I hope one day they get to hear about this and it gives them a happy warm feeling inside

Cathy YesCT

@GossiTheDog this would be easier for folks to understand if you added alt text with the words in the images.

clam

@GossiTheDog John Connor going back in time to shitpost harder and undermine the AI revolution.

Gilgwath

@GossiTheDog Glad to seet that so fare, the BS in BS out axiom still holds. Until that is not the case any longer, I'll postpone the whlle p setting my pants on fire and run in circls in a panic 😁

Richard

@GossiTheDog The fact that this was so easily traceable to one specific post should raise plagiarism and copyright infringement hackles. Even if you could trust LLM output to be true (which you clearly cannot), you can't use it in any real context for fear of unknowingly stealing someone else's work and getting caught or sued for it.

Kaye

@GossiTheDog What could possibly go wrong with asking an automatic Redditor for advice

Fahri Reza

@GossiTheDog the April Fools posts waiting to be discovered

neverbeaten

@GossiTheDog
GIGO.
These are all trained on unfiltered datasets, thus compromising all outputs.
You have to start with the presumption that every AI output is garbage unless proven otherwise by Actual Intelligence. If it happens to be right, it's a lucky coincidence, not by design.

Tristan Harward

@GossiTheDog yet again proving its a really good sounding word model, and nothing even resembling intelligence (which exactly matches its nature).

IanMoore3000

@GossiTheDog I am unable to replicate this result.

Tallawk

@GossiTheDog Google Search is now a gullible 6-year old my how far we've come

Go Up