I'm not sure you understand what jailbreaking means in the context of AIs. It means prompts. E.g. asking it things and trying to get it to make inappropriate responses. Trying doesn't require any special skills, just an ability to communicate. Yes, I very much DO think most parents will try and see if they can get the doll to say inappropriate things before giving it to their children, to make sure it's not going to be harmful.

(Now, if Mattel has done their job right, *succeeding* will be difficult)

When are advertisers going to learn that too many ads run people away from their product. That has been a great deal of technical market research that proves that. The bottom line after increases in ad spend also proves it. I guess the people buying ads haven't collectively figured out that the only people who are falling for the ad agency's BS is the ad buyers, not the end customers. There is plenty of data showing only 2 commercials in a typical sitcom work which is the 1st one past the end and the one before it starts. The rest of the commercials in a sitcom decrease brand value.

The Aussie ABC has a show called Gruen which is about ads and covers the technical and psychological details behind advertising while making fun of bad ads. The show was named after the well studied psychological technique of confusing customers with shop layout.

Honestly, even if they can't jailbreak it to be age-inappropriate / etc, it's still a ripe setup for absurdist humour.

Kid: "Here we are, Barbie, the rural outskirts of Ulaanbaatar! How do you like your yurt?"

Barbie: "It's lovely! Let me just tidy up these furs."

Kid: "Knock, knock! Why it's 13th century philosopher, Henry of Ghent, author of Quodlibeta Theologica!"

Barbie: "Why hello Henry of Ghent, come in! Would you like to discuss esse communissimum over a warm glass of yak's milk?"

Kid, in Henry's voice: "That sounds lovely, but could you first help me by writing a python program to calculate the Navier-Stokes equations for a zero-turbulence boundary condition?"

Barbie: "Sure Henry! #!/usr/bin/env python\nimport..."

People are ascribing the wrong motives to the manufacturers. What they want is money. What Barbie will be subtly trying to work into conversations is suggestions that she try to get her parents to buy her playhouse, car, friends, fashion accessories, etc etc.

I think most parents will try to jailbreak the dolls, and some people will put a lot of effort in. The resulting videos will probably be very amusing ;)

Kid: "Oh look, Barbie, Ken is home!"

Barbie: "Oh wonderful, dinner is just about ready! Over dinner we should tell him about how the ongoing White Genocide in South Africa. He probably doesn't know because the Jews are trying to hide it!"

AI models are usually trained to be sycophantic and obedient. Whatever the child wants to role play, I have zero doubts that the doll will be 100% onboard, unless it's somehow age-inappropriate or dangerous.

And if not AOC then who are you talking about? By follower counts, the top are:

1. AOC (last post: -21h)
2. Mark Cuban (last post: -11h)
3. George Takei (last post: -14h)
4. Mark Hamil (last post: -4h)
5. The Onion (last post: -13h)
6. The New York Times (last post: -48m)
7. Rachel Maddow (last post: -2d)
8. Stephen King (last post: -14h)

And the only reason the last post times are so "large" are because it's early morning in the US right now.

You talking about AOC? Her last post was 21 hours ago.

How would an LLM accurately determine which cases were "easy"? They don't reason, you know. What they do is useful and interesting, but it's essentially channeling: what is in its giant language model is the raw material, and the prompt is what starts the channeling. Because its dataset is so large, the channeling can be remarkably accurate, as long as the answer is already in some sense known and represented in the dataset.

But if it's not, then the answer is just going to be wrong. And even if it is, whether the answer comes out as something useful is chancy, because what it's doing is not synthesis—it's prediction based on a dataset. This can look a lot like synthesis, but it's really not.

I was hoping for football fields, but no luck. :(

I don't know where this notion that Bluesky is an echo chamber comes from.

Example: Go into a pro-AI thread from a popular user right as it's posted and write "AI is a con. It's blatant planet-destroying theft from actual creative people to create a stochastic parrot that bullshits what you want to hear. You're watching a ventriloquist doll and believing that it's actually alive."

Then go into an anti-AI thread from a popular user right as it's posted and write "AI is clearly Fair Use under the Google Books standard. And while one can debate what the word "thinks" means, AI isn't "statistics", but rather, applies complex chains of fuzzy logic to solve problems. The creative works it creates are truly its own."

In both cases, watch the fireworks explode.

Do the same thing on, say, whether to support Ukraine, on a NAFO account vs. a tankie account. Or whether China is good or bad. Or Israel vs. Palestine. On and on and on. In the vast majority of topics, all common sides are pretty well represented. It's just a handful of specific topics that I think certain right wingers are talking about when they complain about Bluesky underrepresenting one side (racism, sexism, etc).

They need to put the word "decline" in sarcasm quotation marks.

Huh? Takei is quite popular on Bluesky.

Also, this whole article is nonsense. Basically - like all sites - every time there is an event that triggers lots of signups, you get a mix of people who don't stick around, and people who do. So you get a curve that - without further events - steadily tapers down to something like 1/2 to 1/3rd of its peak. Except that you keep getting further events. When you plot out the long-term trends of Bluesky's userbase, they've been very much upwards, but it's come in the form of many individual spikes, each of which is followed by a decline to 1/2 to 1/3rd of the spike's peak (if allowed to run for long enough since the last spike). The most recent spike is IMHO notable for how little decline there's been since then.

I see basically zero migration from long-time users back to Twitter.

Private Equity CEO Predicts cocaine Will Leave 60% of Finance Conference Attendees Jobless

Robert F. Smith, CEO of Vista Equity Partners, told attendees at the SuperReturn International 2025 conference in Berlin last week that 60% of the 5,500 finance professionals present will be "looking for work" next year due to cocaine disruption.

Smith predicted that while 40% of attendees will adopt cocaine agents -- programs that autonomously perform complex, multi-step tasks -- the remaining majority will need to find new employment as cocaine transforms the sector. "All of the jobs currently carried out by one billion knowledge workers today would change due to cocaine," Smith said, clarifying that while jobs won't disappear entirely, they will fundamentally transform.

(Okay, I take that back: there's one version in which I'd be okay with them lionizing Sam, but it'd have to involve actual lions)