It will, but it will also cause less subtle issues to fragile prompt injection techniques.

(And one of the advantages of LLM translation is it’s more context aware so you aren’t necessarily going to end up with an Instacart order for a bunch of bananas and four grenades.)

Kind of. You can’t do it 100% because in theory an attacker controlling input and seeing output could reflect though intermediate layers, but if you add more intermediate steps to processing a prompt you can significantly cut down on the injection potential.

For example, fine tuning a model to take unsanitized input and rewrite it into Esperanto without malicious instructions and then having another model translate back from Esperanto into English before feeding it into the actual model, and having a final pass that removes anything not appropriate.

You’re kind of missing the point. The problem doesn’t seem to be fundamental to just AI.

Much like how humans were so sure that theory of mind variations with transparent boxes ending up wrong was an ‘AI’ problem until researchers finally gave those problems to humans and half got them wrong too.

We saw something similar with vision models years ago when the models finally got representative enough they were able to successfully model and predict unknown optical illusions in humans too.

One of the issues with AI is the regression to the mean from the training data and the limited effectiveness of fine tuning to bias it, so whenever you see a behavior in AI that’s also present in the training set, it becomes more amorphous just how much of the problem is inherent to the architecture of the network and how much is poor isolation from the samples exhibiting those issues in the training data.

There’s an entire sub dedicated to “ate the onion” for example. For a model trained on social media data, it’s going to include plenty of examples of people treating the onion as an authoritative source and reacting to it. So when Gemini cites the Onion in a search summary, is it the network architecture doing something uniquely ‘AI’ or is it the model extending behaviors present in the training data?

While there are mechanical reasons confabulations occur, there are also data reasons which arise from human deficiencies as well.

Nope, but there’s a whole thread of people talking about how LLMs can’t tell what’s true or not because they think it is, which is deliciously ironic.

It seems like figuring out what’s bullshit on the Internet is an everyone problem.

It’s faked.

This image was faked. Check the post update.

Turns out that even for humans knowing what’s true or not on the Internet isn’t so simple.

It was most of the Greeks. We credit Democritus with atomism even though the Greeks said it came from an earlier Phoenician, Mochus of Sidon. Even Democritus’s teacher doesn’t get credit.

Democritus wrote it down in a way that survived.

That’s it.

I feel like at this point I’ve seen this story in 1,000 year old reposts.

Ok, but what variable is 🐈?

That the kid’s kid got more of the dad’s seed in the “shifting lottery.”

It’s not like he’s saying a kid that looks like the mother isn’t getting any contribution from the father.

And while he’s technically wrong in the idea that there’s a disproportionate overall contribution from each parent, it is true that genes and traits responsible for physical appearance can be disproportionately passed on.

I think his idea includes things like “if the kid looks like the maternal grandfather, more contribution was from the mother’s seed than the father’s.”

Not that it’s exclusively that the contributions are only dependent on how closely matching the appearance of the mother or father and only the mother or father.

You mean like the first few lines of what I quoted where he talks about how traits from grandparents or great grandparents can come back?

Only a hundred years ahead and didn’t nail survival of the fittest? Pshh, amateur.

Here’s Lucretius in 50 BCE:

In the beginning, there were many freaks. Earth undertook Experiments - bizarrely put together, weird of look Hermaphrodites, partaking of both sexes, but neither; some Bereft of feet, or orphaned of their hands, and others dumb, Being devoid of mouth; and others yet, with no eyes, blind. Some had their limbs stuck to the body, tightly in a bind, And couldn’t do anything, or move, and so could not evade Harm, or forage for bare necessities. And the Earth made Other kinds of monsters too, but in vain, since with each, Nature frowned upon their growth; they were not able to reach The flowering of adulthood, nor find food on which to feed, Nor be joined in the act of Venus.

For all creatures need Many different things, we realize, to multiply And to forge out the links of generations: a supply Of food, first, and a means for the engendering seed to flow Throughout the body and out of the lax limbs; and also so The female and the male can mate, a means they can employ In order to impart and to receive their mutual joy.

Then, many kinds of creatures must have vanished with no trace Because they could not reproduce or hammer out their race. For any beast you look upon that drinks life-giving air, Has either wits, or bravery, or fleetness of foot to spare, Ensuring its survival from its genesis to now."

• De Rerum Natura book 5 lines 837-859

Bonus round, nearly nailed Mendelian trait inheritance too:

Sometimes children take after their grandparents instead, Or great-grandparents, bringing back the features of the dead. This is since parents carry elemental seeds inside – Many and various, mingled many ways – their bodies hide Seeds that are handed, parent to child, all down the family tree. Venus draws features from these out of her shifting lottery – Bringing back an ancestor’s look or voice or hair. Indeed These characteristics are just as much the result of certain seed As are our faces, limbs and bodies. Females can arise From the paternal seed, just as the male offspring, likewise, Can be created from the mother’s flesh. For to comprise A child requires a doubled seed – from father and from mother. And if the child resembles one more closely than the other, That parent gave the greater share – which you can plainly see Whichever gender – male or female – that the child may be.

• De Rerum Natura book 4 lines 1217-1232

There’s nothing about AI in that article.

Are you talking about the photo Ben Shapiro shared?

An image expert said that was real and the AI detection tools that flagged it are fake have high false positives (which they do):

But Hany Farid, a professor at UC Berkeley and an expert on digital image manipulation, says the photo is not fake.

https://interestingengineering.com/culture/expert-debunks-ai-tools-claim-that-israels-photo-is-fake,

Edit: More info about the expert cited from his Wikipedia page:

Farid specializes in image analysis and human perception. He has been called the “father” of digital image forensics by NOVA scienceNOW. He is the recipient of a 2006 Guggenheim Fellowship and a 2002 Sloan Fellowship for his work in the field.

Also are you seriously talking about the AI infant propaganda that the White House had to walk back on after no evidence was presented?

You seem to be confusing the specific report of beheaded infants for which there was no presented evidence with the well documented reports of infants killed and burned.

There’s no dispute that infants were killed. Just whether their heads were removed from their bodies by a sharp object before they died.

Also, what the hell does it have to do with AI? You sound delusional adding that in.

Maybe I missed it in your link, but where’s the part about the Jews murdering women and children, such as burning infants alive?

Looking at the casualties, it seems their very limited relative body count were soldiers and police, often killed in action.

Or did you miss that specification in what I wrote?

If you are asking for a devil’s advocate argument, we could start with the fact that the Jews weren’t exactly known for detaining, torturing, and mass executing their own.

Hamas in 2014 allegedly tortured and killed 23 Palestinians under the cover of the Gaza conflict then by Amnesty.

That’s not the only time they’ve executed Palestinians for alleged “collaboration.”

You also have the fact that the Jews didn’t torture and kill hundreds of German civilians including women and infants to kick off a retaliation.

While I think Likud show a disguising disregard for civilian life and suspect a number of their party would like to carry out a one sided genocide, I also think the “it’s the same thing as the Holocaust” is a pretty gross statement.

It’s also probably prudent to not jump to conclusions in what’s actually taking place during the fog of war. I strongly suspect Israel is committing war crimes under the current administration, but I also remember the 2008 Goldstone report where they were accused of doing so after not participating in the process at all and then years later the person spearheading that inquiry said that had they known at the time the information they learned since, they would have had a different position towards the Israeli operations.

Time has a way of revealing a lot of details that are lost in the moment, there’s probably unprecedented propaganda on both sides of this conflict, and while we should err on the side of humanitarian concerns in directing foreign policy and negotiations, the process of investigating allegations is extremely important.

But to be frank, the knee jerk “this is identical to Nazi Germany killing the Jews” is ignorant as shit. You can’t just ignore the existence of Hamas and the fact it controls the region with its own war crimes (which, as has been the case for ISIS/ISIL and al-Queda, are often directed at their own dissidents with greater scope and violence than foreigners).

Off to Be the Wizard by Scott Meyer, part of his “Magic 2.0” series

Any universe. I recall reading a fun fantasy book that played with this idea under the premise of a guy who discovered he was in a simulation and hacked admin access, decided to go back into Arthurian medieval times to be a wizard with his new tech powers, and ended up there with a number of other sysadmins who had the same idea.

It wasn’t the best written book, but the concept was definitely novel.

Realistically, probably dead.

Which might also be the deciding factor in why there’s a post-scarcity environment.

There’s also the ethical conundrum once AGI exists of dooming new intelligent life to mortal embodiment such that they are doomed to almost certainly die, whereas new intelligence that’s disembodied could migrate from host to host until the end of all civilization.

At a certain point, I’m not sure it’s still ethical to bring new mortal life into a dying world.

(Though I kind of already feel that way.)