- 2 Posts
- 131 Comments
151·5 days agoThey are only doing one segment of users at a time and I’m guessing their metrics are telling them 90% of those users don’t play ball and just stop visiting. I’m guessing they might go back on it in a bit.
It’s a terrible prompt too. They just spammed it with “make the most money possible”. Complex tasks need complex harnesses and multiple agents.
Yes, it exasperates the problem but that’s my point, it isn’t the problem. Trump is literally investing in coal plants. Expecting our energy consumption to stay flat forever is not a solution.
You see the same argument spring up with evs. Evs aren’t the ones burning coal even if that’s what the oil companies were pushing a year ago.
We need a strong grid built on renewables and it should have been started a long time ago.
Don’t fall for this bullshit. The cunt in charge is literally closing down renewable projects left and right.
The constant bribes and corruption are what is slowing down the shift. This is scapegoating.
We need regulations for datacenters and all, and it isn’t helping, but this is transparent propaganda.
I’m okay with using AI to find child abuse as long as we have human verification at the end. It’s more of a privacy issue for me, like how google runs their detection models in private drives and then autobans, very easy to get false positives I imagine.
They mostly are interchangeable for most people but almost none of the llm weights are properly open source.
They don’t always give the training scripts and never the actual data so it isn’t reproducible. The data isn’t really feasible to give. A lot of them release papers at the same time where they explain the training with snippets and how it works, but it’s not the same as receiving a script.
That being said, some get really upset and stress the distinction but it’s a bit silly imo. If it’s free and easily fine tunable, I don’t really see the issue of calling it open source. Giving the data for it would open them up to lawsuits from the copyright mafia and wouldn’t bring much to the table. Training these models is completely out of reach for individual and small to medium sized organisations.
Training a small model was more than a million dollars in compute a year or two ago.
Full text since there isn’t a summary:
Text
One missing underscore in a Skyrim-themed username put an innocent Nova Scotia man in prison for 18 months.
A 2018 child-luring investigation, which began in Madison, Wisconsin, and eventually extended to Halifax, Canada, was based on a false premise.
Police were looking for a man using the Kik messaging service under the name “fus__ro_dah” (two underscores after “fus”), but they accidentally requested records for the username “fus_ro_dah” (one underscore after “fus”). This one-character difference led them not to the perpetrator but to a Canadian man named Brandon Klayme.
(Ars readers may recognize “fus ro dah” as the Unrelenting Force “dragon shout” from The Elder Scrolls V: Skyrim.)
Despite finding no evidence of the crime on his digital devices, Canadian police arrested Klayme in 2020 on child sex abuse charges. He was convicted after a trial in 2023 and sentenced in 2024 to 18 months in prison. He served the full term.
Even after release, Klayme continued to fight his conviction. In the process of preparing his appeal, the username mistake that led to all these years of disruption was finally discovered. On Thursday, the Nova Scotia Court of Appeal overturned Klayme’s conviction, writing: “Mr. Klayme is factually innocent of the offences. He should never have been charged, let alone convicted.” One underscore
The case began in 2018. From August through December of that year, a 12-year-old Wisconsin girl communicated with an adult male through the Kik messaging service. During a check of the girl’s phone, her mother found an “inappropriate” photo of the male and called local police.
The Dane County Sheriff’s Department responded. A deputy took the phone, and the department ran a forensic search on it. The report identified 125 Kik messages between the girl and an adult with the username “fus__ro_dah” (two underscores after “fus”).
To identify this person, the cops contacted Kik, but their subpoena accidentally requested information about the Kik user “fus_ro_dah” (one underscore after “fus”). Kik provided Klayme’s email address in response.
Google records showed that this email address was used to access Google services from an IP address in Canada, so the Dane County investigators turned the case over to Halifax Regional Police. Halifax police took the IP address they had been given to local Internet provider Bell Aliant. Bell connected the IP address to the physical address of their subscriber, Brandon Klayme.
The cops then obtained a search warrant for Klayme’s residence and took phones and laptops from his bedroom. Searching these devices produced no evidence relevant to the case. Nothing linked Klayme to the girl. No intimate images were found. Klayme did have a Kik account, but cops couldn’t even show that he had accessed the service during the period in question.
Still, Klayme was arrested and hit with three charges:
Luring a person under 14 years of age by means of telecommunication Providing sexually explicit material to a child Possession of child pornographyThe case went to trial, where Klayme was found guilty. He then went to prison for 18 months. No one involved, even on the defense side, appears to have noticed the username confusion. The appeal
Klayme continued to press his appeal, but only late in the process did his team figure out what had happened. Here’s how Klayme himself describes the discovery:
In the latter stages of preparing my appeal arguments, it was discovered that the subpoena contained a subtle mistake that changed the course of my life. Instead of requesting details for the username “fus__ro_dah”, the subpoena requested details for the username “fus_ro_dah”. The police officer did not notice that the perpetrator’s Kik username “fus__ro_dah” contains a double underscore. This discrepancy went completely unnoticed at the trial. It was never brought to the attention of the trial judge.After this was pointed out, prosecutors took another look. They agreed that Klayme’s appeal should be allowed. Had the case been properly investigated, the court says now, evidence “would have identified an individual whose first name is Jay whose IP address appears to be in California.”
Given this, the Court of Appeal last week acquitted Klayme on all charges. His sentence (including his ongoing probation) was vacated.
How could such a crucial mistake have been missed for so long? No one seems to know. As the court puts it, “Although the information about the usernames was available at the time of the trial, there is no evidence confirming or explaining how it went unnoticed
He wants regulations to kill the competition. Open source is gonna get fucked in the coming months. The whole media campaign has been specifically for this.
71·9 days agoThe writing styles are actually very different imo.
301·12 days agoWasn’t even a deadly disease. I’m all for gen editing since it has potential to save lives but killing a 6 year okd because you would rather risk it than have an intellectually disabled child seems a bit intense.
They paid like 800k for it so this very much feels like “I’m rich enough not to have to deal with the shame my child will bring me”. They should of just accepted her, what she had was basically down syndrome.
That’s a step away from showing ID. It’s not okay because your local (government owned) library does it. Even wirhout much info, it still becomes easier to dox if you have the person’s location.
It would also be insanely complicated. How many libraries would need to be contacted and taught how to use some kind of software for this. It would be a mess.
“I’m your man for the job. Believe me, I know kids inside and out.”
This guy at the interview probably.
Yes, but taking away the LLM won’t solve anything imo. It’s hard to know if there are more deaths or if it’s just the media pushing it because of what it’s attached too, but I’m guessing someone that falls deep into their delusions because of an LLM would have found something else to do the same with. It’s the system (both the government and probably her immediate support system) that’s letting her down.
It’s kind of muddy, especially when it comes to kids. This case seems like clear cut religion cult delusions though. Hard to pretend the LLM had more impact than her upbringing, and no one talks about restricting religion so it just comes off as another anti-ai token gesture.
It’s like the case of the kid with Daenerys. Everyone wants to talk about the GoT bot but nobody wants to talk about the bullying, the parents who knew about it and let him grow reclusive anyways, or the same parents who left a gun in easy reach.
We don’t regulate things because crazy people read too much into it and lose it, else religion, books and videogames would be. Your claim of AI being different because it has the word intelligence in it is pure bull, specially when video games do have AI lol.
Keep being pedantic about it I guess.
I don’t think the bot was really trying to manipulate but just larping. They are trained on a lot of rpg and make believe data, if you repeat something enough times, it stops saying no and goes into pretend mode, it’s not actually malicious.
That being said, it would be very easy to target people that aren’t mentally stable and then make it worse. I wouldn’t be surprised if it comes out in 20 years that it was the CIA mucking about and causing these suicides or something. This is an insanely convenient way to make sleeper agents.
You’re using your miscomprehension of the words meaning and ignoring his valid point with a trite phrase that means nothing.
Worse, you skipped over video games which does have AI.
The internet in general was probably too dangerous for this person, shall we ban it next?
I find it ridiculous that they are both suing. One of them probably got a better life out of switch.
I don’t know if this link will work, if not you can find the case by going by name and search for parish.
It has a lot of quotes from the convo.
We could probably save more people by restricting religion than LLMs. Whatever real life support system she had, it was probably encouraging her delusions just as much as the bot was.
Are they actually asking for phone numbers now?