Artificial Intelligence (AI) in General

[bill_g]

[Foggy]

"It threw itself down a flight of stairs" my ass. They didn't get the sensors or software right, when stairs are a known hazard. The very idea that "the robot committed suicide" is just silly. This is what I mean when I said a lot of really stupid things are going to be called AI until we actually do get artificial intelligence.

They're not even pretending this robot was an AI, but it takes intention and emotional distress to commit suicide, so they are ascribing AI characteristics to a dumb robot the sneaky way, by saying it deliberately threw itself down the stairs. No, it didn't decide to end its runtime (that's how machine intelligence ends).

Edit: I have to block the top of the stairs in my house when Rochester, my robot vacuum cleaner that Luke talked me into buying, is policing the area. I have a short piece of 2" x 4" that I place across the top of the stairs, and then I never really have to worry about whether Rochester is experiencing some suicidal ideation or sumpin'. I'm really not educated or trained in dealing with suicide, I'd be terrible at figuring out what Rochester's emotional triggers are and how to deal with them. Maybe we can send some 2 by 4s to South Korea, and they can stop worrying about robot seppuku.

[Suranis]

The only way a Robot commits suicide is if it is told to by the instructions given by the various programs in its head.

Anyway xposting

https://vxtwitter.com/1/status/1810284303923720428

Ron Filipkowski (@RonFilipkowski)
A former FL police officer now living in Moscow is creating fake stories using AI while posing as local US newspapers. His fake story that Zelensky’s wife bought a $4 million Bugatti was spread by millions of pro-Russia US social media accounts. https://www.bbc.com/news/articles/c72ver6172do

[bill_g]

Sounds like Ukraine should post pictures of Putin at the beach with his male lover.

[Foggy]

Is a $4,000,000 car a hundred times better than a $40,000 car?

[bill_g]

That was in Euros, and the exchange rate is currently $1US=.92EU.

[neonzx]

Is a $4,000,000 car a hundred times better than a $40,000 car?

Well it looks way cooler.

[Suranis]

https://vxtwitter.com/1/status/1810913528083271746

JerryRigEverything (@ZacksJerryRig)

The United States Justice Department just squashed nearly 1000 of those Russian AI chat bots here on @X today. Glad someone's on top of it. Elons just been retweeting them. https://www.justice.gov/opa/pr/justice- ... r-partners

https://vxtwitter.com/1/status/1810819103390593470

Jack’s House (@FluteMagician)
1) NAFO Fellas, there’s some things mentioned in this article that we were seeing on here for quite a long time now. I was especially interested in how these bot accounts posed as “cryptocurrency” accounts, which we all saw. And Russia Today is involved. Let’s keep going.

[RTH10260]

aged by a few months

Linwei Ding was a Google software engineer. He was also a prolific thief of trade secrets, say prosecutors.
U.S. officials say some of America’s most prominent tech firms have had their virtual pockets picked by Chinese corporate spies and intelligence agencies.

April 6, 2024, 1:00 PM GMT+2
By Ken Dilanian and Anna Schecter.

Though he lived in Silicon Valley, Linwei Ding spent months at a time in his native China, according to court papers.

Nothing unusual about that — except that he was supposed to be working full time as a software engineer in Google’s San Francisco-area offices.

Court records say he had others badge him into Google buildings, making it appear as if he were coming to work. In fact, prosecutors say, he was marketing himself to Chinese companies as an expert in artificial intelligence — while stealing 500 files containing some of Google’s most important AI secrets.

Ding, whose home was searched by the FBI days before prosecutors say he was to board a one-way flight to China, was arrested in March and is now facing federal felony charges. He has pleaded not guilty. His case illustrates what American officials say is an ongoing nightmare for U.S. economic and national security: Some of America’s most prominent tech firms have had their virtual pockets picked by Chinese corporate spies and intelligence agencies.

Days after the Ding case was announced, prosecutors charged the owners of a Chinese company with conspiring to steal battery secrets from Tesla. This week, a government cybersecurity board ripped Microsoft for an “inadequate security culture” and a “cascade … of avoidable errors” that allowed Chinese intelligence hackers to compromise the company’s email software and gain access to the accounts of the U.S. commerce secretary.

In February, the Justice Department charged a Chinese engineer with stealing missile-tracking technology from a company owned by aerospace giant Boeing. Last year, prosecutors accused a Chinese national of stealing Apple’s self-driving car technology and fleeing to China.

“China’s multipronged assault on our national and economic security make it the defining threat of our generation,” FBI Director Christopher Wray told Congress recently. “Today, and literally every day, they’re actively attacking our economic security — engaging in wholesale theft of our innovation and our personal and corporate data.”



https://www.nbcnews.com/tech/innovation ... rcna146623

[SamsOvar]

AI is definitely transforming the way we do things! I’ve been using ai tools for sales to streamline my marketing efforts, and it’s been pretty amazing. For instance, I used an AI tool to analyze customer data and automatically generate personalized emails, which saved me hours of work. It even helped with scheduling and optimizing social media posts.

I’ve found it’s not just about making tasks easier but also about getting better results with less effort. If you’re considering diving into AI, experimenting with these tools can really boost productivity and efficiency.

[Foggy]

I installed Gemini (Goog's AI) on my phone, but I'm not finding it terribly useful yet. Maybe it's me.

I looked at the Google Pickle 9 phones that were just released, and they're loaded with AI this and AI that. They are selling you what they want you to want, not what consumers really want. I honestly have very little need in my life of creating an image of a hippo with a bazooka riding a unicycle in a thunderstorm, but perhaps there are people for whom that's a priority.

I'm stickin' with my Pickle 5a. It still works great. I get 3 spam texts or more every day of my life from various pathetic beggars in the Democratic Party, including a lot from people who wouldn't be my representatives even if they won, and all of them go immediately into my spam folder. I never see them unless I go looking.

That's all I really want from my phone - a good camera and call-and-text screening. Chef's kiss on my Pickle!

[northland10]

Hidden Content

This board requires you to be registered and logged-in to view hidden content.

[RTH10260]

Man Charged in AI-Generated Music Fraud on Spotify and Apple Music

Reporter, Infosecurity Magazine
9 Sep 2024

In a first criminal case involving AI-generated music, a man residing in North Carolina was charged with stealing royalties by using AI to generate fake songs and fake listeners on streaming platforms.

In a September 4 filing, US Attorney Damian Williams charged Michael Smith, 52, for creating a scheme that allowed him to fraudulently obtain $10m in royalty payments from music streaming platforms.

Smith is accused of producing hundreds of thousands of songs with AI, which he published on several streaming platforms and fraudulently streamed using automated accounts commonly known as bots. The targeted platforms included Amazon Music, Apple Music, Spotify, and YouTube Music.



https://www.infosecurity-magazine.com/n ... ic-scheme/

[Foggy]

That's genius, and he has $10M to pay lawyers to show it isn't really a crime.

[RTH10260]

Three Mile Island coming back online to power Microsoft’s growing AI operations

By Jack Aylmer (Anchor), William Jackson (Producer), Ian Kennedy (Lead Video Editor)
Yesterday

The nuclear industry is experiencing a resurgence, driven in part by growing demand for clean energy. Constellation Energy’s decision to revive the Three Mile Island nuclear plant in Pennsylvania is a significant development in this trend.

The plant, which shut down in 2019 due to economic pressures that made it uncompetitive with cheaper natural gas and renewables, is getting a second chance.

Constellation plans to bring the Unit 1 reactor back online in 2028, pending approval from the Nuclear Regulatory Commission. The company is investing $1.6 billion to restart the reactor — a testament to the growing recognition of nuclear power’s role in reducing carbon emissions and meeting increasing demand for electricity

The deal is also a significant step forward for Microsoft, which has committed to powering its growing artificial intelligence operations with clean energy. The 20-year agreement with Constellation will provide 835 megawatts of clean electricity, enough to power a substantial chunk of Microsoft’s data center operations.



https://san.com/cc/three-mile-island-co ... perations/

[tek]

wait woah wut...
TMI unit 1 is fifty years old.

[Foggy]

The 20-year agreement with Constellation will provide 835 megawatts of clean electricity, enough to power a substantial chunk of Microsoft’s data center operations.

Electricity must be transmitted, and some is lost in transmission. The power from this plant isn't going to Redmond, Washington. Is Microsoft going to relocate all its operations to be near 3MI?

[tek]

Maybe there is a nearby datacenter deal in the works?
AWS plopped down a datacenter next to an operating nuke somewhere in PA, and they have a deal to buy the power before it hits the grid and gets transport charges..

[Foggy]

Here's the first really useful thing AI has shown me.

Artificial Intelligence Uncovers 300 New Nazca Lines, Including 'Knife-Wielding Orca'
It took nearly a century to discover a total of 430 figurative Nazca geoglyphs, which offer significant insights into ancient cultures at Nazca Pampa.

https://www.ndtv.com/science/artificial ... ca-6643270

I'm not saying AI is evil (though it can obviously be put to evil purposes), I'm just saying it's not real AI until it has genuine emotions. Until then, I need a new frame for talking about the amazing stuff they're doing in the (fake) name of AI.

But this is cool. I expect ever more amazing things from the technology.

[RTH10260]

Microsoft-Owned LinkedIn Using People's Data To Train Artificial Intelligence Models

by Tyler Durden
Saturday, Sep 21, 2024 - 06:50 PM
Authored by Katabella Roberts via The Epoch Times (emphasis ours),

Professional networking platform LinkedIn has confirmed that it automatically uses personal user data to train artificial intelligence (AI) models without first informing its members.

The California-headquartered company said in a Sept. 18 blog post that it has updated the privacy policy element of its terms of service to include language clarifying how it uses the information shared with it “to develop the products and services of LinkedIn and its affiliates, including by training AI models used for content generation (‘generative AI’) and through security and safety measures.”

The platform said that there is an opt-out setting for members when it comes to using their data for generative AI training.

LinkedIn is owned by Microsoft, which has invested heavily in OpenAI, the developer behind ChatGPT. According to the FAQ section of the platform’s website, the AI models used to power generative AI features may be trained by LinkedIn or another provider, such as Microsoft’s Azure OpenAI service.

The networking site said it uses generative AI for features such as its writing assistant and for suggesting posts or messages.

Personal data such as user posts, usage information, inputs and resulting outputs, language preferences, and any feedback they may provide is among the data processed and used to train AI, LinkedIn said.

When LinkedIn trains generative AI models, it seeks to “minimize personal data in the data sets” used to train them, including by using privacy-enhancing technologies that redact or remove personal data from the training dataset, the company said.

LinkedIn said the updates to its terms of service will go into effect on Nov. 20.

LinkedIn added that it does not currently train content-generating AI models on data from members located in the European Union, European Economic Area, Switzerland, and the United Kingdom.

“If you live in these regions, we and our affiliates will not use your personal data or content on LinkedIn to train or fine-tune generative AI models for content creation without further notice,” the company said.

Opting Out

LinkedIn users in other locations who wish to opt out of allowing the platform to use their data for AI training can visit the “data for generative AI improvement” member setting and set it to “off.”

According to LinkedIn, opting out means that the platform and its affiliates “won’t use your personal data or content on LinkedIn to train models going forward, but does not affect training that has already taken place.”

Meanwhile, Meta, the parent company of Facebook and Instagram, announced earlier this month that it will resume training AI models using public content shared by adults on Facebook and Instagram in the UK over the coming months.



https://www.zerohedge.com/political/mic ... nce-models

[northland10]

I'm eagerly awaiting the fall arrival of Pumpkin Spice AI.

[p0rtia]

[RTH10260]

AI-generated college admissions essays sound male, privileged

By Tom Fleischman, Cornell Chronicle
October 2, 2024

In an examination of thousands of human-written college admissions essays and those generated by AI, researchers found that the AI-generated essays are most similar to essays authored by students who are males, with higher socioeconomic status and higher levels of social privilege.

The AI-generated writing is also less varied than that written by humans.

“We wanted to find out what these patterns that we see in human-written essays look like in a ChatGPT world,” said AJ Alvero, assistant research professor in the Department of Information Science in the Cornell Ann S. Bowers College of Computing and Information Science. “If there is the strong connection with human writing and identity, how does that compare in AI-written essays?”

Rene Kizilcec, associate professor of information science in Cornell Bowers CIS is a co-author of “Large Language Models, Social Demography, and Hegemony: Comparing Authorship in Human and Synthetic Text,” published Sept. 27 in the Journal of Big Data.

Alvero and Anthony Lising Antonio, associate professor of education at the Stanford University Graduate School of Education, are co-corresponding authors.

This research stemmed from Alvero’s dissertation work at Stanford University. Part of his research involved an analysis of approximately 800,000 college admission essays written from 2015-17 by prospective students in the University of California system.

“We consistently found that there was a strong connection between the profiles of the applicants – their test scores, their demographic information, even the high schools they were applying from – and their admissions essays,” Alvero said. “The relationship was so strong that we were consistently able to predict an applicant’s SAT score, within about 120 points.

“The ways that we speak can encode and contain information about our past and who we are,” he said, “and it’s very similar in writing, at least with personal statements.”

For this work, Alvero and the team compared the writing style of more than 150,000 college admissions essays, submitted to both the University of California system and an engineering program at an elite East Coast private university, with a set of more than 25,000 essays generated with GPT-3.5 and GPT-4 prompted to respond to the same essay questions as the human applicants.


https://news.cornell.edu/stories/2024/1 ... privileged

[bill_g]

Microsoft-Owned LinkedIn Using People's Data To Train Artificial Intelligence Models

by Tyler Durden
Saturday, Sep 21, 2024 - 06:50 PM
Authored by Katabella Roberts via The Epoch Times (emphasis ours),

Professional networking platform LinkedIn has confirmed that it automatically uses personal user data to train artificial intelligence (AI) models without first informing its members.
https://www.zerohedge.com/political/mic ... nce-models

Tyler Durden is a fictional character from the movie "Fight Club", and The Epoch Times is an ultra right-wing evangelical Christian pub I wouldn't trust to pour tea without questioning it's motive.

[keith]

Methinks someone just ran up against a case study for the text books of the future...

[tweet][/tweet]