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

[Dakota Tebaldi]

Should that be called the Star Trek Data problem?

I'm leaning more toward the "Galaxy Quest Thermian problem".

 

[Erich Templar]

In a video conference with my boss today he was going through some code that he was working on, and had a co-pilot-like thing installed. It kept popping up with 'helpful' suggestions. I had flash-backs of Clippy...

 

[CronoCloud Creeggan]

In a video conference with my boss today he was going through some code that he was working on, and had a co-pilot-like thing installed. It kept popping up with 'helpful' suggestions. I had flash-backs of Clippy...

I see you're writing a post about an experience with AI, would you like some AI help with that.

 

[Dakota Tebaldi]

TIL

 

[Dakota Tebaldi]

In a video conference with my boss today he was going through some code that he was working on, and had a co-pilot-like thing installed. It kept popping up with 'helpful' suggestions. I had flash-backs of Clippy...

At least whenever Clippy popped up one of his options was always "don't help me anymore with this".

 

[Veritable Quandry]

FAQ: The “Snake Fight” Portion of Your Thesis Defense

March 13 2022 4/24/23 5/8/24 CALENDAR 2027

www.mcsweeneys.net

At Arkansas State you usually got a cottonmouth.

 

[Free]

Google promised a better search experience — now it’s telling us to put glue on our pizza

Google’s new AI Overview feature is telling users a whole bunch of wrong answers.

www.theverge.com

Very smart newsmedia...

Google AI said to put glue in pizza — so I made a pizza with glue and ate it

Google's AI Overviews suggested adding glue to pizza sauce. I tried it. And ate it. Here's what all this tells us about the future of the web.

www.businessinsider.com

 

[Argent Stonecutter]

Regwall

 

[Dakota Tebaldi]

I don't know if I really am that interested in the opinion of someone who actually made pizza with glue and ate it for the sake of a headline, knowing that it was idiot behavior and doing it anyway?

But anyway yeah that aside. I hate so much about this. It's astounding that Google is doubling down on a product that is so self-evidently not ready for public use yet, that they're perfectly okay with their Magic Answer Machine giving just completely, even dangerously wrong answers because "it only does that sometimes, not like ALL the time".

But you know, I hate the idea of this "feature" even if it worked absolutely perfectly. Maybe it's weird of me to care about this at all, but this thing marks a complete transformation of Google Search's entire function and purpose.

Google was a "search engine". It indexed websites, and when you used it you were asking for a list of websites that potentially had the answer to the thing you were looking for or asking about, and it gave you a list of links, and then you clicked on those links and went and explored those websites.

This is not that. AI-enhanced "search" isn't search at all, it just scrapes the information from other websites, rephrases it, and then presents it at the top of your "search results" so that you can just read an answer to your presumed question and be satisfied and never actually leave Google, except maybe to go to sponsored product pages whenever your search query can be somehow tied to a thing someone is selling for money. That's going to strangle informational websites that aren't Google, and I kind of think that's part of the point. Google is trying to reinvent itself into an everything-aggregator and reinterpreter, and it's trying to get users on board with that by spinning just freaking using the internet and browsing websites as an inherent inconvenience, a problem that they have invented a solution for.

 

[Bartholomew Gallacher]

AI is creating jobs, or: Google scrambles to manually remove weird AI answers in search

Google scrambles to manually remove weird AI answers in search

Google’s AI Overview launch showcases that the race for AI domination is perilous.

www.theverge.com

 

[Govi]

AI is creating jobs, or: Google scrambles to manually remove weird AI answers in search

Google scrambles to manually remove weird AI answers in search

Google’s AI Overview launch showcases that the race for AI domination is perilous.

www.theverge.com

Since that is the current state of Google's AI efforts, it's clear why they were so conservative about launching it (until Microsoft's Bing stole the march on them). Clearly, publicly scrambling to catch up is not working to Google's benefit.

 

[Bartholomew Gallacher]

 

[Noodles]

AI is creating jobs, or: Google scrambles to manually remove weird AI answers in search

Google scrambles to manually remove weird AI answers in search

Google’s AI Overview launch showcases that the race for AI domination is perilous.

www.theverge.com

I can see it now.

Everytime one shows up, they just add an if/else statement to "rerun the results" or just return something like "Google is overloaded right now sorry."

 

[Free]

Everytime one shows up, they just add an if/else statement to "rerun the results" or just return something like "Google is overloaded right now sorry."

GOTO DUCKDUCKGO

 

[Bartholomew Gallacher]

Tons of more Google fun, like: smoke 2-3 cigarettes during pregnancy. Doctor's order.

 

[Free]

 

[Dakota Tebaldi]

Tons of more Google fun, like: smoke 2-3 cigarettes during pregnancy. Doctor's order.

LOL the hilarious thing about this is that Google can't back down. It's invested so many resources into its very bad chatbot that it can't just take it down and say it needs some work, they have to really commit to just saying that it's fabulous right now and that giving very bad and sometimes dangerously wrong answers isn't a drawback but just a charming quirk that users will have to deal with.

 

[Bartholomew Gallacher]

Google is really begging for tons of class action law suits with that type of behavior right now. Question is not if, but when this will start.

 

[Argent Stonecutter]

They're still not bloody AI.

 

[Dakota Tebaldi]

They're still not bloody AI.

On a certain level I think even its most hardcore stans realize this which is why sooooo much of the messaging and advertising around "AI" products is based on just outright deception. Faked demos of fake products and it's okay because it's totally gonna be able to do that for real by this time next year, trust me bro.