The anti-AI thread

Page 22 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.
Toggle sidebar Toggle sidebar
IronWing

IronWing

No Lifer
Jul 20, 2001
73,516
35,206
136
I don't mind AI generated stuff that is presented as such. As of last week my birder group has been inundated with bots posting AI bird photos. The noise destroys communication, which is the intent.
 
  • Like
Reactions: Jon-T and marees
mikeymikec

mikeymikec

Lifer
May 19, 2011
21,470
16,698
136
Red Squirrel said:
I hate that it also says "monitors". That was my fear, that they start bringing "smart TV" garbage to monitors too. Ugh. I just want my screen to be a screen.
Click to expand...

How are they going to give a monitor an Internet connection in a manner not easily avoidable?

I suspect this is simply going to be an opt-in for "technology enthusiasts".

On a separate note - I did a file search in Win11 recently (on a customer's PC) and the Explorer status bar informed me that my search was being powered by AI which may lead to incorrect results being delivered, along with a feedback link. Has anyone else seen this?
 
Kaido

Kaido

Elite Member & Kitchen Overlord
Feb 14, 2004
52,214
7,550
136
gizmodo.com

Microsoft Forced Copilot Into Your Life. Now It Can Accidentally Drain Your Bank Account

The agent has your wallet.
gizmodo.com gizmodo.com

Per Microsoft, Copilot Checkout will allow users to have a conversation with Copilot about products and make a decision to purchase right from the chat. When they decide to buy, an AI agent powered by Copilot navigates to the retailer and completes the transaction for them. Copilot Checkout, which will be available through Copilot.com, will start with partnerships with Etsy, PayPal, Shopify, and Stripe. Microsoft says it’ll ramp up available retailers this month.

Retailers who operate using PayPal or Stripe can apply to become an eligible Copilot Checkout merchant, but Shopify sellers will automatically be enrolled in the program and have to manually opt out if they don’t want to participate
Click to expand...

 
  • Wow
Reactions: Red Squirrel
mikeymikec

mikeymikec

Lifer
May 19, 2011
21,470
16,698
136
NVIDIA CEO has the sads because people are saying mean things about AI and therefore denying him the chance to build another 27 swimming pools. Also, it'll be the naysayers' fault when the AI bubble pops, all this poor little puppy needed to survive was a few GDPs, was that too much to ask.

slashdot.org

Nvidia CEO Jensen Huang Says AI Doomerism Has 'Done a Lot of Damage' - Slashdot

Nvidia CEO Jensen Huang "said one of his biggest takeaways from 2025 was 'the battle of narratives' over the future of AI development between those who see doom on the horizon and the optimists," reports Business Insider. Huang did acknowledge that "it's too simplistic" to entirely dismiss...
slashdot.org
 
Last edited:
  • Haha
  • Like
Reactions: KompuKare, Racan, Saylick and 1 other person
nakedfrog

nakedfrog

No Lifer
Apr 3, 2001
63,369
19,748
136
mikeymikec said:
NVIDIA CEO has the sads because people are saying mean things about AI and therefore denying him the chance to build another 27 swimming pools. Also, it'll be the naysayers' fault when the AI bubble pops, all this poor little puppy needed to survive was a few GDPs, was that too much to ask.

slashdot.org

Nvidia CEO Jensen Huang Says AI Doomerism Has 'Done a Lot of Damage' - Slashdot

Nvidia CEO Jensen Huang "said one of his biggest takeaways from 2025 was 'the battle of narratives' over the future of AI development between those who see doom on the horizon and the optimists," reports Business Insider. Huang did acknowledge that "it's too simplistic" to entirely dismiss...
slashdot.org
Click to expand...
" I think we're scaring people from making the investments in AI that makes it safer, more functional, more productive, and more useful to society"
MF, people invested over $200 billion in AI last year. How about you invest in a bit of GFY?
 
  • Like
Reactions: KompuKare, Racan, Saylick and 1 other person
IronWing

IronWing

No Lifer
Jul 20, 2001
73,516
35,206
136
My company just pushed another MS Office update that keeps us from turning off Copilot. I'm tempted to use the shit out of it. "You want shit products? Here's your shit products."

There is an option to tell Copilot to only be annoying if it thinks something is really important. This tells me that Copilot is real-time analyzing everything I type. I wonder how the c suite folks feel about MS having access to everything every employee types as they type it?
 
  • Like
Reactions: marees
nakedfrog

nakedfrog

No Lifer
Apr 3, 2001
63,369
19,748
136
IronWing said:
My company just pushed another MS Office update that keeps us from turning off Copilot. I'm tempted to use the shit out of it. "You want shit products? Here's your shit products."

There is an option to tell Copilot to only be annoying if it thinks something is really important. This tells me that Copilot is real-time analyzing everything I type. I wonder how the c suite folks feel about MS having access to everything every employee types as they type it?
Click to expand...
"Well, it's good for my stock portfolio, so I'll allow it"
 
Kaido

Kaido

Elite Member & Kitchen Overlord
Feb 14, 2004
52,214
7,550
136
nakedfrog said:
I guess it's a "fun" preview of what we're in for.
Click to expand...

imo AI slop will be the 2026 "pandemic". Wild goose chases, "swatting", false accusations, fabricated evidence, companies pushing data-collection overreach, environmental abuses from datacenters, electricity prices jacked up, computer hardware shortages, etc.

www.the-independent.com

Fake AI-generated videos and photos about Venezuela receive millions of views

Within hours of Donald Trump’s announcement that Venezuela’s president had been captured, social media was flooded with AI-generated media. Bryony Gooch reports
www.the-independent.com www.the-independent.com

A big part of the problem is simply human nature:

1. People believe what they see

2. People tend to only read headlines

3. People are already pre-disposed to certain worldviews (left or right politics, conspiracy believers, etc.)

Hard to fight fake news when people WANT to buy into whatever reality is being peddled to them ¯\_(ツ)_/¯

6abc.com

'No one verified the evidence': Woman says AI-generated deepfake text sent her to jail

Melissa Sims said her nightmare began after she called police during an argument with her boyfriend, when she said he allegedly ransacked her home.
6abc.com 6abc.com
 
Last edited:
  • Like
Reactions: KompuKare
Red Squirrel

Red Squirrel

No Lifer
May 24, 2003
71,164
14,016
126
www.anyf.ca
I saw that a while back, pretty interesting phenomenon but also not surprising. I always feel like using AI is cheating but do find myself using it and it's 10x faster and better than googling. I don't know what to think about it, I don't necessarily like it but I also can't ignore a resource that can help me.

Will be interesting to see what happens long term, because AI relies on non AI data to train. When people are just posting AI responses and AI trains on that, it will cause some kind of feedback loop where no new data is actually learned. I'm sure this is something even the AI companies are going to have trouble with.
 
Kaido

Kaido

Elite Member & Kitchen Overlord
Feb 14, 2004
52,214
7,550
136
Red Squirrel said:
I saw that a while back, pretty interesting phenomenon but also not surprising. I always feel like using AI is cheating but do find myself using it and it's 10x faster and better than googling. I don't know what to think about it, I don't necessarily like it but I also can't ignore a resource that can help me.

Will be interesting to see what happens long term, because AI relies on non AI data to train. When people are just posting AI responses and AI trains on that, it will cause some kind of feedback loop where no new data is actually learned. I'm sure this is something even the AI companies are going to have trouble with.
Click to expand...

I was wondering about that too:

1. Gathering existing data

2. Vetting existing data

3. Creating NEW data

Some stuff it can solve (ex. coding) through simulations, but it needs data for health, human emotions, physical effects in the universe, etc. Will be interesting to see how they prevent a data-rot spiral, as no human can validate ALL of the data in the AI databases!
 
Kaido

Kaido

Elite Member & Kitchen Overlord
Feb 14, 2004
52,214
7,550
136
This is pretty cool, but I'm putting it in this thread because, you know...yikes

This is off-the shelf Nano Banana Pro & Kling 2.6:

www.instagram.com

Instagram

Create an account or log in to Instagram - Share what you're into with the people who get you.
www.instagram.com www.instagram.com

https://imgur.com/a/this
 
Last edited:
Steltek

Steltek

Diamond Member
Mar 29, 2001
3,448
1,190
136
Well, so much for Microslop security:

Reprompt: The Single-Click Microsoft Copilot Attack that Silently Steals Your Personal Data

What is worse is that the method is likely applicable to most AI chatbots.

And, remember how all the AI companies claimed in court that their products were learning "patterns" and were not storing the actual contents of the utilized training material?

A Stanford research paper that just dropped pretty much proves beyond a doubt that the AI companies are purposely lying through their teeth to the courts in all the legal filings they have made related to the issue. Either that, or they truly don't understand how what they are creating actually works. I don't know which possibility is worse......

Abstract paper is here:

Extracting books from production language models

Many unresolved legal questions over LLMs and copyright center on memorization: whether specific training data have been encoded in the model's weights during training, and whether those memorized data can be extracted in the model's outputs. While many believe that LLMs do not memorize much of...
arxiv.org arxiv.org

futurism.com

Researchers Just Found Something That Could Shake the AI Industry to Its Core

Researchers found compelling evidence that AI models are actually copying copyrighted data, not "learning" from it.
futurism.com futurism.com

 
  • Like
Reactions: KompuKare, KMFJD and mikeymikec
Kaido

Kaido

Elite Member & Kitchen Overlord
Feb 14, 2004
52,214
7,550
136
Steltek said:
And, remember how all the AI companies claimed in court that their products were learning "patterns" and were not storing the actual contents of the utilized training material?
Click to expand...

"Harrison Ford as Rambo"

Uh-huh :colbert:

1768678722208.png
 
  • Love
Reactions: Steltek
You must log in or register to reply here.

TRENDING THREADS

COMPANY
RESOURCES
FOLLOW
Top Bottom