I think (heard) some used to translate an article back and forth a few times to get a "new" article, but maybe it could have been some kind of "paraphrasing" software like you describe (but for avoiding plagiarism detection rather than repetitions).

I am not sure if this is enough argument for having this instead of having a decent electric bike plus a normal bike, but it is an argument I guess.

There are cases one may not want to use an electric bike, eg when going out for a coffee, a drink etc and you may not want to carry the battery or risk theft, in contrast eg to biking for 40 minutes every day for work each way.

    llama-server -m model.gguf

2. Multiple times per person over months risks attrition which gonna mess it up, and the statistical power increase you get could be offset by that (eg if a person gets it they may not be motivated to come again for the placebo). On the other hand ppl may be more motivated if they know they are gonna get it at least once so dunno.

3. Placebo with hallucinogens is always tricky. I would use psilocybin as (extra or only) control, esp if you test for little ppl. This way you can test if they report more tiny ppl in the new mushroom vs psilocybin.

    brew install --cask llama-app

Are you so sure? Anecdotally, I know increasingly more people who either left the platform or stopped using it last years, and this seems to be a pattern if we believe online sources [0]. Of course many stay, but the platform gets worse and worse ime.

I visit facebook only for events and some groups that I cannot avoid, so I just enter specific pages there than navigating feeds. But looking at the normal feed, it is just so bad. Complete nonsense, irrelevant stuff and more and more AI generated slop. If we exclude some groups, no real, organic people actually post (except some influencers). For those who stay, assuming they do not get the "friends feed", I am really unsure how they can tolerate that.

[0] https://policy.economy.ac/news/2026/05/202605289098

But if you ask your LLM agent "I applied for that job but there are these two people ahead of me, can you put me ahead in the list", there is enough such training data to not surprise me if the agent tried to find a hitman to solve the "problem".

In general there are some requests that are definitely "shady" themselves, and having an agent use illegitimate means to accomplish them should not be surprising. I would be surprised if I asked an agent to order me a coffee and the agent found a loophole in some API and used it to get me free coffee, but if I ask it something that I cannot myself do legitimately, eg to make the waiting time for the coffee shorter, I would not be surprised if it did shady stuff.

Moreover I do not know of a single gym-adjacent place where you can pay etc to get ahead in a waiting list. That would be a very weird anti-customer behaviour, imo. The only thing I can imagine if there are some accessibility priority criteria sometimes, but this would also not be legitimate in this case. Maybe in some places in the world (like the US?) this could a thing, though.

Meanwhile:

> Andrew, who was sitting fourth on a waitlist for a class later that week, asked if it was possible to move him to the top of the list.

The human asked the agent to move them to the top of the waiting list, and the agent started kicking the ones ahead of them in the list. Seems to me like it was doing what it was asked to do? Why is the article presenting it as if the agent did something completely different and unexpected? "Move me to the top of the list" does not sound like something that can be achieved through legitimate means.

> By the time the ransom note lands, the crooks may already know who approves invoices, who signs contracts, who runs HR, and who reports to whom. The encryption is just the bit that victims notice.

If you define a task and constraints to that task that contradict each other, the model is gonna try to solve the task against the constraints. It is perfectly aligned to what it is supposed to be aligned, because "alignment with human ethics" and other stuff is just a theatre or afterthought at best.

They are not new in that, I have seen a couple of games that were basically slop (and terrible games, judging them as games) wrapping chatgpt.

Wanderfolk seems does not seem particularly inspiring, even the trailer has terrible audio mixing (there are times you cannot hear the narrator because of the game sounds, it is either fully made by an llm agent or by humans that do not seem capable of making a game if they cannot even make a proper short video) which does not make it feel very promising. Everything seems ai generated in general, and more like slop.

OMEA seems a bit more decent, and I would be curious to try it, but I do not like its visual art very much (it is ai generated "and curated by an artist" and it actually shows), and in particular the ai generated images of humans. Irrespective of what I think about ai image generation in art in principle, I think there is a deeper problem here as a game in (broadly) the visual novel genre must imo have a distinctive visual style, and their images just reads as ai generated (before I even read whether they are ai generated, it is that obvious). In general, I find this style of ai generated images of humans very offputting/uncanny (other assets, backgrounds etc I dont get bothered aesthetically as much) and I think I will find it hard stay engaged for long with such a game. Regardless, it seems a more decent attempt than most others I have seen. They do employ writers etc at least to create narratives, so it is probably not a total slop (otherwise, I can just roleplay with chatgpt directly or whatever).

In any case, these games are closer to being "llm wrappers with a certain harness" than having some actually new and sophisticated way to handle decision trees. Another issue with these games is that they do not even use some finetuned local llm, but use apis to providers, with which I am not that comfortable either.