kombookcha

Agreed. I also feel like there is a world of difference between the driver deliberately assuming control at the last second because they notice that an accident is about to happen, and the car itself yielding control unprompted because it thinks an accident is about to happen.

The former is to be expected. The latter seems likely to potentially make an already dangerous situation worse by suddenly throwing the controls to an inattentive driver at a critical moment. It seems like it would be much safer for the autopilot to continue doing its best while sounding a loud alarm to make it clear that something dangerous is happening.

Maybe you gotta do daily standups with all your own Agents and then have both you and them do standup with the rest of the team's Agents to touch base, sync up and loop back or whatever.

The second somebody manages this, I will rush out and throw my credit card at their store so hard it embeds itself in the counter like a ninja star.

Stay hopeful, maybe Nu-Agile will become annoying in new and interesting ways!

Yeah, I agree with you that this is probably the best course of action in terms of minimal investment of time and minimal exposure. And in general, you get a lot further in life by trying to be amicable as your default stance! I want to be kind, and most other people do too!

The thing that makes me wary about recommending carrot over stick here, is that it might long term enable thoughtless behaviour from the people deploying the bot, by offloading their shoddy work into a shadow time-tax on a bunch of unseen external kindly people. But if deploying pushy or rude robots means you risk a nonzero number of their victims shoving something into the gears to get rid of it, then that incurs a cost on the owner of the bot instead.

Of course, it may also just lead to bad actors making more combative or sneaky bots to discourage this. There aren't really any purely good options yet.

One can imagine an agentic highwayman demanding access to your data, first politely, and then 'or else'.

I don't want to be flippant, but why is anyone else responsible for play-acting with somebody's uninvited puppet?

I get that you could probably finagle a way to get it to fuck off by play-acting with it, and that this would probably be the easiest short term fix, but I don't think that's a reasonable expectation to have of anyone.

Prompt injecting a hostile piece of software that's hassling you uninvited is an annoying imposition for the owner, but the bot itself being let loose is already an annoying imposition for everyone else. It's not anyone elses job to clean up your messy agent experiment, or to put it neatly back on its shelf.

I don't remember if it's ever explicitly stated, but I certainly took that to be the implication. Maybe it's actually funnier if they're doing it pro bono, now I am suddenly unsure.

He was! Frito Pendejo (Joe's lawyer) and the cop Beef Supreme are better examples. Or of course the wrestler-president Camacho whose middle name is literally Mountain Dew.

Commenting to save this for later.

Thank you for sharing this.

The really funny thing is that both feel equally plausible from the vibe he puts out.

One of my favourite LinkedIn cranks is a consultant guy who suggested an addendum to Einstein that "Has the potential to impact the future: E=mc²+AI". Because, so he says, "This equation highlights the potential for AI to unlock new forms of energy, enhance scientific discoveries and revolutionize various fields such as healthcare, transportation and technology".

It got a lot of traction with the algorithm due to lots of comments to the effect of "What??", but every time I pass by the screenshot in my photoroll, I have a little giggle about it.

Oh boy, if it's powered by AI /AND/ the blockchain, you know it's gonna be legit ;)

Eleven years seems like a very long time to be a Philly street dog - kinda makes you wonder if it wasn't adopted by somebody in the interrim before ending up with the girl somehow.

Exactly. The price to actually do this is simply not one the US is willing to pay.

I think this is the elephant in the roomt - in terms of quantifiable goals, Iran is winning this thing. I think they're going to want to punish the US and Israel to an extent where they will be reluctant to feel this particular sting again, and they want to assert their ability to control the strait. And it's working! They're clearly demonstrating that the US cannot simply decide when this is over and dictate terms, because Iran can pinch off an important vein of global commerce and probably sustain that pressure for far longer than it can be tolerated by other economies.

They've already gotten one concession in terms of this temporary sanctions relief, even as Trump frames it as a domestic emergency measure and repeatedly declares total victory each day of the conflict. They also got him to back off on targeting their power plants by promising to retaliate in kind against the power infrastructure of US aligned states in range.

I think the US has the ability to beat Iran in a fight, but it does not have the preparation or the resolve to do so at this time, because this is some halfcocked nonsense plan with amorphous goals that they thought would be over in a week.

Oh, I misread your post - that's fair!

The rejection is based on the dishonesty of explicitly committing to standard A and then knowingly violating it, not on LLM use as such. I think that's pretty fair, considering that everyone could have just chosen B if they wanted to.

I think the fundamental problem is that humans use language to refer to things and constructs that exist and have various relationships with eachother in meatspace, whereas LLMs use words solely as things that exist in relation to other words. That's inherently lossy if you're trying to make it fetch and regurgitate information encoded in the former format.

While the ability to interface with a computer program in plain language is the really interesting thing here IMO, it also comes with a number of problems baked in that are worse than person-to-person transfers of text-speech.

Your monkey brain is actually quite good at figuring out if other monkeys are bullshitting you and what they mean, because you can make use of a vast number of small cues and unconscious tells in what they say and how they say it - even in writing. With an LLM, you cannot do this because it will always have the same confident can-do zeal with everything you ask it for.

A boss at my former job would constantly generate AI images and send them to the graphics guy for him to make 'adjustments' and no matter how many times the graphics guy explained that these images are much more trouble to clean up than for him to make a new one from scratch by hand, he seemed unwilling to understand and kept spamming slop at him.

I really felt for the guy the first time I was in a meeting and somebody had generated their own project roadmap recommendations. This type of behaviour introduces so much noise and time waste in the system, I would love to know how it shapes up next to the benefits.

Don't even get me started on people AI generating personal farewell notes for retiring coworkers or whatnot.

LLMs by definition cannot deduce, because they cannot not know or think. There's guard rails to try to make it more correct than wrong, but ultimately it's about which words seem like they would fit when coming after your words.

It's a neat trick, but the mind wants to ascribe meaning and reason to words that sound meaningful and reasonable, but these words do not come from a thinking mind with intent and interiority. It would be much more interesting if they did, but when and if that does happen, it won't be from an LLM as we know them today.

That sounds insufferable and I'm sorry your coworker and your bosses have gotten this hypnotized. There are actual use cases, but this type of productivity cargo cult drives me up the wall.

At least if they were into filling the office with magic crystals, it would be decorative and easily ignored. This is just forcing people to spend a bunch of tokens in a dull ritual to make the line go up.

They make it their business to replace resignees with loyalists. In terms of generating friction for the regime, it makes more sense to make your stand in office and make them fire you, but that carries a lot of personal/career costs too. Which I suspect is why people choose walking away instead.

Preemptive compliance does nobody any good.

I read somewhere that Australia has 10-14 days of gasoline supply - I have no idea how many countries run this tight and exposed, but it seems to be surprisingly widespread - it's wild how the global economy runs on just-in-time shipping of energy imports.

It's unfortunate but the marketing spin has transcended into public discourse, to the point where people take these wild aspirational claims about LLMs for granted. As opposed to what they are - sales pitches.

It doesn't help that there is now so much capital tied up in these products whether directly or indirectly that a lot of people and organizations find themselves unwilling to face the reality of what the products can and cannot do.

What a wonderful read.

This is already sort of in Jurassic Park 3, which features a 3D-printed velociraptor larynx that can be blown into to reproduce vocalizations remniscent of the velociraptors in the movie. At the end of the movie Alan Grant blows in it, confusing a pack of velociraptors.

There's a specific call that the movie has established as velociraptors in distress calling for help/backup from the pack, and Alan attempted to reproduce it when cornered at the beach to basically bluff/intimidate the pack into avoiding a fight. It coincided with the arrival of a helicopter, and the combined effect made the raptors run away.

Not explicitly a language as such, but specific calls with understandable, relatively complex meaning to the dinosaurs, that can be (sorta) understood and leveraged by humans.

This would probably still be way better than the status quo, because it would introduce a higher barrier for entry and more friction into the slop delivery system and give the moderation and spam detection a chance to catch up. Much easier to detect and delete 5000 laundered slop images than 500.000.

Guarding your heart with elegant nonsense you don't really mean is a classic defensive posture, and probably is directly impeding their ability to be present in emotionally intense (and often difficult) situations. It reminded me of this:

>There is a scene in the opening of Into the Abyss. Werner Herzog is interviewing a Reverend who in fifteen minutes will go in to be with a boy as the boy is led to the gurney to be executed by injection.

>The Reverend is talking about how the Lord works in mysterious ways, and so on—it is exactly the type of conversation you want to avoid. It is very ChatGPT. It is the Reverend repeating things he’s said before—words that protect him, that allow him to perform the role of Reverend, instead of being what he is: a man named Lopez, who will soon have to watch a boy die.

>At one point, the Reverend, as a part of a monologue about the beauty of God’s creation, mentions that he sometimes meditates on the beauty of the squirrels he sees on the golf course. Herzog, standing in a graveyard with nameless crosses, says, with mad Bavarian seriousness, “Please describe an encounter with a squirrel.”

>Lopez is a bit surprised by the question, but he takes it in a playful spirit—his voice lifts, joyously. He starts to talk faster. (This is where the conversation shifts into the type you want.) He is no longer saying versions of things he has said before, he’s not protecting himself, he’s just there.

>From that point on, it takes about ten seconds before he’s crying.

>In interviews, Herzog likes to mention this conversation to explain his craft. “But how on earth did you know to say that?” says the interviewer. “Were you just trying to say something unexpected to unbalance him?” “No, it was not random”, Herzog says. “I knew I had to say those exact words. Because I know the heart of men.”

https://www.henrikkarlsson.xyz/p/looking-for-alice

Hah, you got me there! I'll try to keep the ratio flipped correctly next time ;)

I read that article recently, so the similarities might not be entirely coincidental. 'JPEG of thought' is gonna stay in my vocabulary for a while.