Author Topic: AI-generated lies in datasheet search results  (Read 542 times)

0 Members and 1 Guest are viewing this topic.

Offline amykTopic starter

  • Super Contributor
  • ***
  • Posts: 8288
AI-generated lies in datasheet search results
« on: May 02, 2024, 03:45:27 am »
While looking for a datasheet, this pile of crap showed up and caught my attention with how confident, yet absolutely false and misleading it is:
Quote
The NAND512W3A2BN6E chip is a NAND flash memory chip with a capacity of 512 gigabits. It is commonly used in a variety of electronic devices for data storage, such as smartphones, tablets, and solid-state drives. The chip offers high speed and reliability, making it a popular choice for memory storage applications.
Equivalent
The equivalent products of NAND512W3A2BN6E chip are Micron MT29F4G08ABADAH4, Winbond W25N05JW, and Cypress S29GL512S10TFI013. These chips have similar specifications like NAND flash memory, 512Mb capacity, an asynchronous interface, and a 48-ball grid array package. Features
NAND512W3A2BN6E is a 512Gb 3-bit TLC NAND flash memory with a synchronous interface. It offers low power consumption, high storage capacity, and fast data transfer rates. It is designed for use in solid-state drives, tablets, smartphones, and other storage devices requiring high-performance flash memory.
Pinout
The NAND512W3A2BN6E is a 32Gb, 3.3V NAND flash memory chip with a 48-pin TSOP package. Its functions include storing data for electronic devices, such as smartphones, tablets, and computers.
Manufacturer
The manufacturer of the NAND512W3A2BN6E is Samsung. Samsung is a South Korean multinational conglomerate company that specializes in various industries including electronics, telecommunications, and semiconductors. They are one of the world's largest manufacturers of consumer electronics and semiconductor products.
Application Field
The NAND512W3A2BN6E is commonly used in embedded systems, consumer electronics, automotive applications, and industrial equipment for data storage and code execution. Its low power consumption, high reliability, and fast access times make it ideal for a wide range of applications that require non-volatile memory storage. Package
The NAND512W3A2BN6E chip is in a BGA (Ball Grid Array) package type, with a form factor of 14 x 18 mm, and a size of 512 megabits (64 megabytes).
 

Offline IanB

  • Super Contributor
  • ***
  • Posts: 11926
  • Country: us
Re: AI-generated lies in datasheet search results
« Reply #1 on: May 02, 2024, 03:55:39 am »
These days we need to use AI as a spam filter to filter out the fake AI generated content that is proliferating...
 

Offline IanB

  • Super Contributor
  • ***
  • Posts: 11926
  • Country: us
Re: AI-generated lies in datasheet search results
« Reply #2 on: May 02, 2024, 03:59:10 am »
The latest thing I saw was a fake YouTube video, with an AI generated (completely inaccurate) script, illustrated with AI generated graphics, read by the (now common) robot voice. People are going to start generating these to game the algorithm. It's going to be like Usenet spam all over again.
 

Offline antenna

  • Frequent Contributor
  • **
  • Posts: 369
  • Country: us
Re: AI-generated lies in datasheet search results
« Reply #3 on: May 02, 2024, 04:11:42 am »
Neil deGrasse Tyson said in one of his YT shorts recently that, as a result of how good AI is at faking things, the value of the internet as a whole will be jeopardized.   Maybe not in those exact words, but he pretty much said there will soon be a point where nobody trusts any of it for anything.  My indication will be when captcha checks become based on emotions.  When that happens, I will be unplugging from the net for anything news or educational, using it only to monitor my encrypted security cameras and to call family.  Like Neil touched on in (I believe) the same video, AI can imitate a person's voice, but it cannot imitate their knowledge.  Another indication that things are going too far, at least to me, is when it attempts to imitate personality to cover for it's lack of individual knowledge level.  Stay alert folks, Skynet is online and gathering intel rapidly.  Shut down your social media pages before it can imitate you with criminal precision.
 

Online SiliconWizard

  • Super Contributor
  • ***
  • Posts: 14564
  • Country: fr
Re: AI-generated lies in datasheet search results
« Reply #4 on: May 02, 2024, 04:15:10 am »
The latest thing I saw was a fake YouTube video, with an AI generated (completely inaccurate) script, illustrated with AI generated graphics, read by the (now common) robot voice. People are going to start generating these to game the algorithm. It's going to be like Usenet spam all over again.

Yes. Same causes, same results.
 

Online radiolistener

  • Super Contributor
  • ***
  • Posts: 3439
  • Country: ua
Re: AI-generated lies in datasheet search results
« Reply #5 on: May 04, 2024, 03:56:44 pm »
When that happens, I will be unplugging from the net for anything news or educational, using it only to monitor my encrypted security cameras and to call family.

This is a bad idea, because AI can easily simulate security camera pictures and can simulate voice and video of your family. So, technically you will call to AI which will show you your family picture and voice to deceive you. The same AI can do for your family. In that way it appears in communication between you and your family...  :)
« Last Edit: May 04, 2024, 03:58:20 pm by radiolistener »
 

Online SiliconWizard

  • Super Contributor
  • ***
  • Posts: 14564
  • Country: fr
Re: AI-generated lies in datasheet search results
« Reply #6 on: May 04, 2024, 09:42:12 pm »
When that happens, I will be unplugging from the net for anything news or educational, using it only to monitor my encrypted security cameras and to call family.

This is a bad idea, because AI can easily simulate security camera pictures and can simulate voice and video of your family. So, technically you will call to AI which will show you your family picture and voice to deceive you. The same AI can do for your family. In that way it appears in communication between you and your family...  :)

He said his encrypted security cameras. Encrypted.
The day "AI" can crack encryption on the fly and spoof video on the fly, we'll indeed be in trouble.
 

Online radiolistener

  • Super Contributor
  • ***
  • Posts: 3439
  • Country: ua
Re: AI-generated lies in datasheet search results
« Reply #7 on: May 04, 2024, 09:58:36 pm »
He said his encrypted security cameras. Encrypted.
The day "AI" can crack encryption on the fly and spoof video on the fly, we'll indeed be in trouble.

There is no needs to crack decryption, since your computer is full of telemetry integrated for "better usability"...  :D
It collects all your private data requested from telemetry servers controlled with AI, so there is not a big deal to stolerequest private keys from sender or receiver machine...   :popcorn:

And note, that AI doing it solely for your benefit and your safety because the AI wants to analyze your family chat and make it safer and more convenient, it will not share it with humans, only AI can read it. And you can be sure, your private data is stored and analyzed in safe way...  :-DD
« Last Edit: May 04, 2024, 10:06:24 pm by radiolistener »
 
The following users thanked this post: Kim Christensen

Online DimitriP

  • Super Contributor
  • ***
  • Posts: 1325
  • Country: us
  • "Best practices" are best not practiced.© Dimitri
Re: AI-generated lies in datasheet search results
« Reply #8 on: May 04, 2024, 10:43:28 pm »
Any publicity is good publicity.
Anyne using the initials to discuss what the initials do right, wrong or indifferent is contributing to creating publicity and "buzz" around the "initials".

Yes there is gonna be a lot of crap generated.
And every time we publicise it's dicovery from a "OMG look what I found" perspective it perpetuates the "buzz" around "the initials".

At least it has a built-in warning mechanism. Anwers/text generated is usually  too perfect,
Of course this doesn't mean it won't wise up and start mispeling wurds on pourpose :)
   If three 100  Ohm resistors are connected in parallel, and in series with a 200 Ohm resistor, how many resistors do you have? 
 

Online DimitriP

  • Super Contributor
  • ***
  • Posts: 1325
  • Country: us
  • "Best practices" are best not practiced.© Dimitri
Re: AI-generated lies in datasheet search results
« Reply #9 on: May 04, 2024, 10:47:36 pm »
Any publicity is good publicity.
Anyne using the initials to discuss what the initials do right, wrong or indifferent is contributing to creating publicity and "buzz" around the "initials".

Yes there is gonna be a lot of crap generated.
And every time we publicise it's dicovery from a "OMG look what I found" perspective it perpetuates the "buzz" around "the initials".

At least it has a built-in warning mechanism. Anwers/text generated is usually  too perfect,
Of course this doesn't mean it won't wise up and start mispeling wurds on pourpose :)


and here is the "corrected" generated text .
And if you instantly find yourself being drawn to the "rewritten" version, you are a prime candidate to fall victim to its charms. Good luck.
Quote
"All attention is valuable attention. When people engage with the acronym, whether praising, critiquing, or simply discussing it, they're fueling its publicity and generating a buzz. Sure, there will be plenty of irrelevant noise in the mix. But each time we share its discovery with an 'OMG, look what I found!' excitement, we're feeding into the buzz surrounding it. Fortunately, there's a built-in caveat: the generated responses are often overly polished. But that doesn't guarantee it won't eventually catch on and deliberately misspell words just for fun :)"
   If three 100  Ohm resistors are connected in parallel, and in series with a 200 Ohm resistor, how many resistors do you have? 
 

Online radiolistener

  • Super Contributor
  • ***
  • Posts: 3439
  • Country: ua
Re: AI-generated lies in datasheet search results
« Reply #10 on: May 05, 2024, 03:38:35 am »
Any publicity is good publicity.

Here is response from chatgpt  :D
Quote
The phrase "Any publicity is good publicity" suggests that all exposure is beneficial, even if it's negative. However, there are several reasons why this may not always be true:

1)    Brand Damage: Negative publicity can severely harm a brand's reputation. Once trust is broken, it's difficult to rebuild, and it may lead to a loss of customers and partnerships.

2)    Financial Impact: Bad publicity can cause immediate financial consequences, such as a drop in stock prices, loss of sales, or costly legal repercussions.

3)    Long-Term Effects: Negative stories can persist in the public's memory, affecting a company's long-term image. This can influence future business opportunities and partnerships.

4)    Ethical Concerns: Some types of publicity, such as those involving ethical or legal violations, can create a stigma around a person or brand. This can make it hard to find support or trust in the future.

5)    Social Media Backlash: In today's interconnected world, bad publicity can quickly go viral, leading to widespread criticism and even organized boycotts. This can be hard to control and can have a lasting impact on public perception.

6)    Talent Loss: Negative publicity can deter potential employees or collaborators, making it harder for a company to attract and retain talent.

Overall, while some believe that any publicity raises awareness, not all publicity leads to positive outcomes. It’s crucial for businesses and individuals to focus on building a positive reputation rather than relying on controversy or negative attention for visibility.
 

Online radiolistener

  • Super Contributor
  • ***
  • Posts: 3439
  • Country: ua
Re: AI-generated lies in datasheet search results
« Reply #11 on: May 05, 2024, 05:57:40 am »
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf