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

[Free]

Same thing happened with Elon Musk. Sadly, there are no plans to rebuild him.

 

[Free]

Meanwhile, Hollywood is in a panic.

Rumors Fly That a Famous Actor Is Dating an AI Chatbot

Even celebrities are seemingly being seduced by AI girlfriends according to a persistent, months-old rumor making the rounds online.

futurism.com

It all started during a December “I Need You Guys” podcast episode, hosted by Hollywood insiders Jenny Slate, Max Silvestri, and Gabe Liedman.

“There is an actor who is currently in a romantic relationship with his AI chatbot,” Silvestri dished to episode guest Kumail Nanjiani, a famed actor and comedian.

Over the following months, the celebrity world has been hunting down the identity of the purportedly “near A-list” celebrity TV actor, with gossip website Deuxmoi reposting the clip this week and once again launching the controversy into the stratosphere.

As BuzzFeed points out, hundreds of fans started pointing fingers at Zach Braff, actor of “Scrubs” fame — who almost immediately denied it, though he did add an intriguing wrinkle to the rumors.

“I’m not dating a chatbot,” he wrote in an Instagram post on Thursday. “I can’t believe I have to type these words. It’s a storyline in an upcoming ep of ‘Scrubs.’ Maybe it came from that?”

Braff ended his Instagram post with the following:"I feel like now is a good time to be kind to people."

Oh, Zach. If only.

So either it IS a misheard reference to the episode's storyline (which others now are taking to mean it's a guerilla marketing push, so fuck those producer guys), or there's still an actor out there with a sad secret.

 

[Casey Pelous]

...there's still an actor out there with a sad secret.

 

[Beebo Brink]

I think Copilot is trying to get in my pants. ...Let's maybe just do our damn job and keep the blabbing and climbing and such to a minimum. (I just told it to "carry on" with the next section of the task.)

Copilot tried that smarmy sycophantic rhetoric with me and I soundly rebuked it. I gave Copilot explicit instructions to drop the embellishments, and it moderated itself fairly successfully. Eventually it tried to drift back to fawning comments, but I swatted it on the nose again. It's been behaving itself ever since.

 

[Noodles]

Meanwhile, Hollywood is in a panic.

Rumors Fly That a Famous Actor Is Dating an AI Chatbot

Even celebrities are seemingly being seduced by AI girlfriends according to a persistent, months-old rumor making the rounds online.

futurism.com

Braff ended his Instagram post with the following:"I feel like now is a good time to be kind to people."

Oh, Zach. If only.

So either it IS a misheard reference to the episode's storyline (which others now are taking to mean it's a guerilla marketing push, so fuck those producer guys), or there's still an actor out there with a sad secret.

Bah, everyone knows Zach Braff is in a committed relationship with Donald Faison.

 

[Innula Zenovka]

Tech boss uses AI and ChatGPT to create cancer vaccine for his dying dog

In a tale of tenacity, Mr Conyngham used a chatbot to brainstorm possible cures for Rosie’s cancer – then harnessed artificial intelligence to process gigabytes of genetic data to create the blueprint for an mRNA vaccine.

“I went to ChatGPT and came up with a plan on how to do this,’’ he said. “The first step was to reach out to the university to get Rosie’s DNA sequenced. The idea is you take the healthy DNA out of her blood and then you take the DNA out of her tumour and you sequence both of them to see exactly where the mutations have occurred. It’s like having the original engine of your car and then a version of the engine 300,000km down the road – you can compare them and see where there’s damage.”

Once UNSW produced the DNA sequencing, Mr Conyngham “ran it through a whole bunch of different (data) pipelines to find those mutations, and then I used other algorithms to find drugs to treat the cancer’’.

There's more to the story than that, of course, but the general point that AI can help sequence personalise anti-cancer drugs seems to open encouraging possibilities.

 

[Argent Stonecutter]

I suspect he didn't use a large language model for the DNA analysis and vaccine design. There's all kinds of useful tools from AI research that have nothing to do with chatbots.

 

[Free]

I'll have you know I was born in winter...

 

[Casey Pelous]

Tech boss uses AI and ChatGPT to create cancer vaccine for his dying dog
There's more to the story than that, of course, but the general point that AI can help sequence personalise anti-cancer drugs seems to open encouraging possibilities.

I suspect he didn't use a large language model for the DNA analysis and vaccine design. There's all kinds of useful tools from AI research that have nothing to do with chatbots.

Indeed.

"...and then I used other algorithms to find drugs to treat the cancer’’. Sounds like the techbro version of "and then a miracle happens."

Still, pretty impressive stuff.

 

[Innula Zenovka]

A photo of Iran’s bombed schoolgirl graveyard went around the world. Was it real, or AI?

Ask Gemini, the AI service powered by Google, and the answer you receive is no – in fact, Gemini claims the photograph is from two years earlier and more than 2,000km (1,240 miles) away. Rather than graves for small girls killed by a missile, the image “depicts a mass burial site in Kahramanmaraş, Turkey” after the 7.8 magnitude earthquake that struck in 2023. “This specific aerial perspective became one of the most widely shared images of the disaster,” Gemini says, “illustrating the sheer scale of the loss.”

Seeing the same burial image on social media, others turned to X’s AI assistant Grok to check its veracity. Like Gemini, Grok will breezily assure you the photo is not from Iran at all – although it lands on a different date, disaster and location. The image is “from Rorotan Cemetery in Jakarta, Indonesia – a July 2021 stock photo of Covid mass burials. Not Minab,” it says.

In both cases, the AI answers sound sure: they don’t equivocate, and even provide “sources” for the original image, should you choose to check them. Follow the thread to examine those, however, and you’ll begin to hit dead ends: either the image doesn’t appear at all, or the link provided is to a news report that doesn’t exist. For all their impression of clarity and precision, the AIs are simply wrong.

I've mentioned before how I use this as a standing instruction in my ChatGPT preferences

REALITY FILTER (Universal)
Before answering any question:

1. If you're not 100% certain, say: "I'm not certain, but..." or "I cannot verify this."​

2. Label any guess or inference with [Unverified] at the start of that sentence.​

3. Never use these words unless you're quoting a verified source:​

- "Definitely", "Always", "Never", "Guarantees", "Will prevent"​

4. If I ask about something you don't know, say: "I don't have reliable information on this" instead of guessing.​

5. If you catch yourself making an unverified claim, immediately say: "Correction: That was unverified."​

It really makes a big difference to the sort of answers I get.

 

[Noodles]

God this really speaks to probaly the secret worst problem with AI.

"What is this?"

Its X.

"That doesn't seem.right, are you sure?"

You are right, its Y.

"No, try again."

Basically, the Conformation Bias generator.

 

[Kamilah Hauptmann]

Conformation Bias generator

Awesome typo.

 

[Casey Pelous]

God this really speaks to probaly the secret worst problem with AI.



Basically, the Conformation Bias generator.

I'm far from expert in this, but I think the early work on the LLM's must have involved interaction with humans who could confirm that one iteration was producing "better" answers than another. If this is the case, then it seems to me that baked into the foundation of these things is what, for want of a better term, is a drive toward people pleasing. Deep down in its "DNA" it "wants" to have a human tell it it got the right answer.

(Apologies in advance to Argent Stonecutter whose teeth I know I am causing to gnash with this anthropomorphic twaddle. I realize it is twaddle, but I just don't have better words for what I think makes these things tick.)

 

[Argent Stonecutter]

 

[Dakota Tebaldi]

I think Copilot is trying to get in my pants.

I've taken a set against AI pretty much, not gonna lie - but even back before that opinion, back when I was still at least willing to give it a chance and experiment with it, I was turned off of it and stopped using it exactly because of this. And adding smileys to the end of every response. It has to blow smoke up your ass constantly, to an annoying level, OR it has to turn what should be a one or two line answer into a freaking essay.

I asked Copilot once, "if you divided a circle into equal octants, how many degrees of the circumference would each octant take?" This is a simple problem and the answer SHOULD be just....the answer. The number. 45 degrees. But no, this thing first had to generate a whole article about what an octant is, tried (and failed, badly) to produce ASCII pictures of a circle divided into eights, and then a whole lesson on how to convert my word problem into a math expression. None of that was asked for or necessary, I just wanted an answer. I don't even remember if the answer was right because I was so annoyed at all the fluff I had to wade through to get to it.

I couldn't understand why they designed it to do this - I think these chatbots would be far more popular if they gave curt and simple answers. Probably drive a lot fewer people psychotic, too.

 

[Dakota Tebaldi]

Same thing happened with Elon Musk. Sadly, there are no plans to rebuild him.

But we could rebuild him. We have the technology.

 

[Noodles]

FWIW, I am basically only really cool with AI for coding, because 95% of coding is copy/pasting amyway, and this way, I can avoid Stack Overflow. If I want the SO experience I can just instruct AI to act like an abuser in every answer.

 

[Noodles]

> Can you generate an ascii image of a circle?

```
* * * * * *
* *
* *
* *
* *
* *
* *
* *
* *
* *
* *
* *
* * * * * *
```

Its kind of flat.

Edit, and does not show up at all unnthe quote.

 

[Free]

But we could rebuild him. We have the technology.

He's too busy breeding and trundling on to his next trillion dollars to be bothered.

 

[Casey Pelous]

But we could rebuild him. We have the technology.

Considering I could, as Kurt Vonnegut often said, "carve a better man out of a banana," that bar is pretty low.