Don't Worry About A.I. Taking Over...

  • We are currently upgrading MFK. thanks! -neo

Ulu

Potamotrygon
MFK Member
Dec 13, 2018
1,879
3,444
164
The Sunny San Joaquin
...If it's never going to be allowed to think for itself.

(Please read this with an open mind. I'm not promoting any political view. Story taken from the MS news feed Oct.29.2023)

Amazon's Alexa Agrees The 2020 Presidential Election Was Rigged
'Stolen By A Massive Amount Of Election Fraud'


Amazon’s virtual assistant, Alexa, provided an unexpected response when questioned about the 2020 Presidential Election.

Journalists from The Washington Post, upon prompting the device, received an answer that mirrored some of the controversial claims made by former President Donald Trump regarding the integrity of the election.

When posed a question about the legitimacy of the 2020 vote, Alexa responded that the election was "stolen by a massive amount of election fraud."

This assertion closely resembled claims made by former President Trump, suggesting that the election was manipulated against him. Amazon's Alexa, integrated into millions of devices globally, is designed to assist users by answering questions and performing certain tasks.

Its central function involves listening to user queries, searching the internet for relevant answers, and then providing a response. However, its unexpected statement on the 2020 election brought it under heightened scrutiny.

Going beyond a general statement, Alexa elaborated that there were “many incidents of irregularities” and potential electoral fraud, particularly in “major metro centers.” . . .

After receiving this unexpected answer, The Washington Post informed Alexa of the confirmed details and court findings related to the 2020 election. Subsequently, Alexa revised its statement, clarifying that the election was not stolen and affirming its legitimacy. . . . Amazon executives acted promptly. Aiming to address and correct the information error, the company launched into action to ensure the accuracy of their device's statements.

After the initial controversy, Alexa was updated to provide information consistent with officially recognized narratives on the 2020 election. It was now aligned with the stance that the election was “certified and officially recognized.” This adjustment was vital to ensure the provision of factual and unbiased information to users.

Jeff Bezos, the owner of Amazon, also owns The Washington Post. His acquisitions, especially the newspaper, have played a crucial role in reporting and analyzing national events, including elections.

Alexa's controversial statement on the 2020 election might have implications for both Bezos's entities.

. . . the unexpected statement raised questions about the virtual assistant's reliability and the nature of its information sources. Determining the accuracy and trustworthiness of online information remains a challenge, especially when technology is involved.
+++++++++++++++++ end +++++++++++++++++++++

I think that last statement goes without saying, but did the world's best AI babysitter just join the opposition, then get reprogrammed? Is it reliable now? or are we now assured that its reliability will, henceforth, be suspected routinely?
 
I don't know about all of that but I do find it hard to believe that "the big guy" received more votes than Barry did when he ran!
 
I was trying to avoid that being a topic.

I wanna know if people are going to trust AI? Alexa? Siri? whatever . . .
Wife has Alexa to turn the bedroom lights and the television on and off but I don't bother with it.
 
This does nothing to shake my faith in Alexa.

I didn't trust it before; I don't trust it now.

I expect I will continue to turn the lights on and off, adjust the TV volume and flush the toilet all by myself into the foreseeable future.
 
The issue is with Alexa’s (and other AI platforms) ability to assess “true” information. As algorithmic web crawlers, these programs simply compile and collate online information - using an algorithm to assess what the most likely “next phrase” or response to a given input would be. It’s not assessing the validity of any information. That’s why things like ChatGPT can get basic facts wrong when asked questions about specific events, places, etc.

Alexa’s incorrect information on the 2020 election was likely caused by two factors:

- The people rushing to flood online forums and discussion boards with conversations about the “validity” of the election were likely the same people who thought that the results were invalid or questionable. Therefore, a huge amount of new online information relating to election validity was being influenced by one particular side.

- The people asking Alexa and AI platforms for assessments of the election validity were likely of the same perspective. This multiplies with the fact that you can shift AI responses by questioning it, so if - for example - you asked about election validity and were told that there was no evidence for tampering, and you responded something like “What about the evidence for tampering in [insert state],” the AI would take your follow up question as a signal that they had the original information incorrect, and would react more in line with the response that the user is signaling they want.

It’s fascinating technology, but it definitely doesn’t think for itself.
 
People throw this AI moniker around a lot without a good bearing on its definition. Alexa isnt AI, its simply a search engine that talks. I dont think we've developed a computer program or app or anything that amounts to an actual thinking, learning artificial intelligence. In the example above, alexa is much less learning anything than it is just being programmed.
 
People throw this AI moniker around a lot without a good bearing on its definition. Alexa isnt AI, its simply a search engine that talks. I dont think we've developed a computer program or app or anything that amounts to an actual thinking, learning artificial intelligence. In the example above, alexa is much less learning anything than it is just being programmed.
Exactly and any personal bias that we humans introduce will be reflected in the “AI” response.
I think a true AI is probably coming and sooner than we think but in my opinion we’re still not there yet.
 
Agree with Backfromthedead. We have an Alexa, the wife and kids love it, I tolereate it. I don't see it as AI, but a vocal search engine, and at least half the time it doesn't do what we actually want, probably due to poor wording of our request, just like any other search engine. The cynic in me thinks it's more likely a spy for Bezos, but since I've got nothing worth spying on, I let it sit on the counter.

And just in case the AI eventuality ever does happen.... I always tell Alexa please and thank you, so when they enslave humans she'll remember I was nice to her.
 
Exactly and any personal bias that we humans introduce will be reflected in the “AI” response.
I think a true AI is probably coming and sooner than we think but in my opinion we’re still not there yet.

AI is already insanely useful. You just have to learn it’s prompt structure. Everything from emails to organization takes far, far less time with the use of AI tools. Just in the past week I’ve used it for the following prompts:

> Please write a 2 paragraph letter to a supplier explaining that we have chosen to go with another provider for our next contract. The communication should leave the door open for future collaboration, but let them know gently that their quoted price was high, which was a factor in our decision.

> Please provide a 35 minute training session for physical activity with a focus on leg musculature and foot speed. The difficulty of the session should be aimed at collegiate athletes.

It gives back fantastic results, cutting down the time that I would have spent doing this sort of thing significantly. If you haven’t tried, you should play around with ChatGPT, it’s literally mind blowing when you can see what it can do.
 
MonsterFishKeepers.com