Nobody cares about "AI" (Chatbot: I disagree.)

[Free]

The Internet Enabled Mass Surveillance. A.I. Will Enable Mass Spying.

The internet enabled mass surveillance. A.I. will enable mass spying.

slate.com

Spying and surveillance are different but related things. If I hired a private detective to spy on you, that detective could hide a bug in your home or car, tap your phone, and listen to what you said. At the end, I would get a report of all the conversations you had and the contents of those conversations. If I hired that same private detective to put you under surveillance, I would get a different report: where you went, whom you talked to, what you purchased, what you did.

Before the internet, putting someone under surveillance was expensive and time-consuming. You had to manually follow someone around, noting where they went, whom they talked to, what they purchased, what they did, and what they read. That world is forever gone. Our phones track our locations. Credit cards track our purchases. Apps track whom we talk to, and e-readers know what we read. Computers collect data about what we’re doing on them, and as both storage and processing have become cheaper, that data is increasingly saved and used. What was manual and individual has become bulk and mass. Surveillance has become the business model of the internet, and there’s no reasonable way for us to opt out of it.

Spying is another matter. It has long been possible to tap someone’s phone or put a bug in their home and/or car, but those things still require someone to listen to and make sense of the conversations. Yes, spyware companies like NSO Group help the government hack into people’s phones, but someone still has to sort through all the conversations. And governments like China could censor social media posts based on particular words or phrases, but that was coarse and easy to bypass. Spying is limited by the need for human labor.

A.I. is about to change that.

 

[Free]

Meta’s new AI image generator was trained on 1.1 billion Instagram and Facebook photos

Imagine with Meta AI" turns prompts into images, trained using public Facebook data.

arstechnica.com

On Wednesday, Meta released a free standalone AI image-generator website, "Imagine with Meta AI," based on its Emu image-synthesis model. Meta used 1.1 billion publicly visible Facebook and Instagram images to train the AI model, which can render a novel image from a written prompt. Previously, Meta's version of this technology—using the same data—was only available in messaging and social networking apps such as Instagram.

 

[Beebo Brink]

Another useful AI tool.

NEEDS MORE BOOM
Bogart and Rains walking away together at the end of Casablanca directed by Michael Bay
FADE IN:
EXT. AIRSTRIP - NIGHT

The runway is bathed in ethereal moonlight, a solitary PLANE SHIMMERS in the distance. A dramatic ORCHESTRAL SCORE swells.

WHIP PAN to CAPTAIN LOUIS RENAULT (Rains), his expression is grave. He flicks a cigarette into the distance. As it lands, a MASSIVE EXPLOSION engulfs the background. Renault barely flinches.

Suddenly, a sleek, BLACK LAMBORGHINI roars onto the scene. The door swings open revealing RICK BLAINE (Bogart). He steps out, suave as ever.

The SOUND of GUNFIRE in the distance. The two men look at each other. They've been through this too many times.

BOGART
(Raises an eyebrow)
Looks like Strasser got a bigger welcome than he anticipated.

Rains chuckles as we CUT TO a MONTAGE of WWII themed TRANSFORMERS wreaking havoc in the distance.

Rains turns to Bogart, a twinkle in his eye.

RAINS
I think this is the start of a beautiful friendship.

Suddenly, a ROBOTIC CAMEL (Yes, a robotic camel) descends from the sky, the ground shakes as it lands. The camel sways as it kneels, inviting them for a ride.

They hop on. The camel shoots off, leaving a blazing trail of fire and sand in its wake.

A DRONE SHOT reveals a crisscross of bullets and explosions in the distance. Our heroes weave through the chaos, unscathed.

Bogart looks at Rains, his face lit by the glow of the explosions.

BOGART
Here's looking at you, kid.

Suddenly, the camel leaps into the air transforming into a FIGHTER JET. The duo flies away, leaving the chaos behind as "AS TIME GOES BY" BLASTS through the speakers.

FADE OUT:

Just as Rains is about to light another cigarette, the lighter EXPLODES. Rains gives a wry smile.

FADE IN:

Casablanca (In flaming letters) from the jet's POV as they fly away.

ROLL CREDITS.

 

[Free]

Another company from Elon Musk. </kidding!>

Humans have to intervene in 70% of orders, says AI drive-thru firm

Presto Automation pitches a talking chatbot which is able to take drive-thru orders with very little human intervention - and 'never forgets to upsell', the firm's website reads.

www.dailymail.co.uk

An artificial intelligence company which works with fast food giants Carl's Jr, Hardee's, Del Taco and Checkers has admitted that human intervention is needed on the majority of orders.

Presto Automation pitches a talking chatbot to take drive-thru orders - which never forgets to upsell'. It hopes to help cope with widespread labor shortages in the sector.

But the firm has said that 'off-site agents' help during more than 70 percent of orders to make sure the AI system does not mess up, according to company filings.

Before AI destroys humanity, it might take out corporations first.

 

[Noodles]

God I can't even ever get basic mobile orders right. Now you want some AI trying to turn my talking into a mobile order that can be screwed up?

 

[Innula Zenovka]

Whoever would have expected it?

 

[GoblinCampFollower]

Whoever would have expected it?

This also makes me think about how Gates said chat bots might be plateauing. They aren't true intelligence, they just spit out content based on input data. I think Machine Learning is to AI what an abacus is to computers. A truly intelligent AI is a long way off. ChatGPT doesn't seem to have a magical solution for this either... they just have more moderation.

 

[Innula Zenovka]

This also makes me think about how Gates said chat bots might be plateauing. They aren't true intelligence, they just spit out content based on input data. I think Machine Learning is to AI what an abacus is to computers. A truly intelligent AI is a long way off.

Exactly. These Large Language Model AIs are stochastic parrots. They're very good at pulling together strings from their huge corpus of reading and turning it into coherent English sentences that look like plausible responses to the questions people ask, but they have no way of assessing the accuracy of their output.

Train your AI on Twitter X and the output will inevitably contain a good deal of nonsense, because a good deal of what's posted on Twitter is nonsense.

Garbage In, Garbage Out, in other words.

Constrain the dataset to mainstream media, and you'd get much cleaner results.

This is, I think, completely different from the AI tools that assist with things like medical diagnostics, which, as I understand it, are good at spotting statistical significant connections between various pieces of patient data and possible diagnoses.

 

[GoblinCampFollower]

Exactly. These Large Language Model AIs are stochastic parrots. They're very good at pulling together strings from their huge corpus of reading and turning it into coherent English sentences that look like plausible responses to the questions people ask, but they have no way of assessing the accuracy of their output.

Train your AI on Twitter X and the output will inevitably contain a good deal of nonsense, because a good deal of what's posted on Twitter is nonsense.

Garbage In, Garbage Out, in other words.

Of course. nothing wrong with the tech, just that it's not the magical oracle some want it to be...

Constrain the dataset to mainstream media, and you'd get much cleaner results.

But the mainstream media tells me things I don't want to hear! It is funny how some of the most gullible fools will brag about how they think they are too smart for the media, but will believe ANYTHING that actually fits their world view...

This is, I think, completely different from the AI tools that assist with things like medical diagnostics, which, as I understand it, are good at spotting statistical significant connections between various pieces of patient data and possible diagnoses.

of course! those AI's are trained by programmers working closely with experts who know what they are doing... that tech is very promising.

 

[Free]

More whoopsie-doodling in the legal realm.

Michael Cohen’s lawyer cited three fake cases in possible AI-fueled screwup

Lawyer David Schwartz must explain why a motion cited "cases that do not exist."

arstechnica.com

A lawyer representing Donald Trump's former attorney Michael Cohen filed a court brief that cited three cases that do not exist, according to a federal judge. The incident is similar to a recent one in which lawyers submitted fake citations originally provided by ChatGPT, but it hasn't yet been confirmed whether Cohen's lawyer also used an AI tool.

"On November 29, 2023, David M. Schwartz, counsel of record for Defendant Michael Cohen, filed a motion for early termination of supervised release," US District Judge Jesse Furman wrote in an order to show cause yesterday. "In the letter brief, Mr. Cohen asserts that, '[a]s recently as 2022, there have been District Court decisions, affirmed by the Second Circuit Court, granting early termination of supervised release.'"

Schwartz's letter brief named "three such examples," citing United States v. Figueroa-Florez, United States v. Ortiz, and United States v. Amato. The brief provided case numbers, summaries, and ruling dates, but Furman concluded that the cases are fake.

"As far as the Court can tell, none of these cases exist," Furman wrote. Furman, a judge in US District Court for the Southern District of New York, ordered Schwartz to provide copies of the three cited decisions by December 19.

Posted here for the headline, as Schwartz has not responded on the how or why the citations were...chosen.

 

[Veritable Quandry]

The article I read did not mention AI, but it was the first thing I thought of.

 

[Noodles]

This also makes me think about how Gates said chat bots might be plateauing. They aren't true intelligence, they just spit out content based on input data. I think Machine Learning is to AI what an abacus is to computers. A truly intelligent AI is a long way off. ChatGPT doesn't seem to have a magical solution for this either... they just have more moderation.

The fact that you have to spend a shitload of time and energy training a brain snapshot for these things is proof enough that they are not "intelligent".

Call me when it can do it in real time.

Call me when it can web search for information it does not know, and get the right asnwer.

 

[Innula Zenovka]

The perils of courtroom AI

Look out for hallucinating claimants, judges are told

rozenberg.substack.com

Evernote link

 

[Free]

These AI-generated news anchors are freaking me out

Channel 1's eerily real newscasters suggest a tumultuous future for online video.

arstechnica.com

 

[Noodles]

Ebay seems to have some new AI stuff when creating listings that makes very wrong guesses for like, weight and category.

 

[Bartholomew Gallacher]

Too bad this friendly AI sales bot is running under supervision.

 

[Noodles]

I wonder how integrated it is with systems. Probably none, but maybe you could tell it, "Schedule the deleivery of a Chevy Tahoe to (an address, like a park or something).

 

[Cristiano]

GPT and other AI models can't analyze an SEC filing, researchers find

The findings from Patronus AI highlight some of the challenges of using AI models within big companies in regulated industries like finance.

www.cnbc.com

 

[Noodles]

I have been playing ith Notebook LLM, where you can feed it material and get results. I fed it a few Second Life PDFs I have.

i asked it "Write an LSL script to open and close a door"

Here is an example of a script that you could use to open and close a door in Second Life:

default
{
touch_start(integer total_number)
{
state open;
}
}


This script will open the door when it is touched. To close the door,
you would need to add an additional state to the script, like so:
default
{
state closed;
touch_start(integer total_number)
{
if (state == closed)
{
state open;
}
else
{
state closed;
}
}
}

This script will toggle the state of the door between open and closed each time it is touched. You can also add animations to the door to make it look like it is actually opening and closing.
Keep in mind that you may need to modify this script to fit your specific needs, such as changing the state names or adding additional functionality.

 

[Erich Templar]

I have been playing ith Notebook LLM, where you can feed it material and get results. I fed it a few Second Life PDFs I have.

i asked it "Write an LSL script to open and close a door"

Interesting. It looks correct, but incorporates one bad practice (if a touch changes state, you should do it touch_end not touch_start).