Jobs Lost To AI

[Irrev-Black]

Gizmodo is apparently in the process of ditching its non-English edition staff, and replacing their work with bot-translated derivations of the English version.

https://www.theverge.com/2023/9/1/23856 ... anslations

[stylofone]

I found this article about about AI to be informative and thought-provoking.

https://www.abc.net.au/news/science/202 ... /102988050

Sitting on my front verandah I offer some of my TLDR type thoughts...

My reaction was that it exposes the limits of AI. All that effort, all that huge corporate wankery that went on just to get a computer to do one thing, win at a game. The human grand master Lee Sedol obviously didn't like losing, same as Kasparov at chess. But in another sense, Lee's capability to feel triumph or disappointment is actually a greater achievement than his game-playing ability, if you want to measure it by the complexity required to emulate it artificially. And on top of that, like all humans he would have countless other skills and abilities which a thousand Deepminds couldn't even begin to emulate. This is why I think the Singularity is bullshit, part of a fantasy beloved of self-indulgent tech bros. As with cryptocurrency, the question of the electricity needed to run all those computers hosting the AIs is already being raised. Maybe Moore's law will close the gap somewhat, but I suspect there will be an AI equivalent of a lightspeed limit which will stop the Singularity from happening, and the way climate change is going, high tech economies are heading for a nasty shock and won't get anywhere near that particular event horizon anyway.

For the nerds, I add that there is a truckload of science fiction on this subject, and I am most familiar with Iain M Banks' oeuvre. It's horrifying that Jeff Bezos is also a fan, he can't have read those books closely, and Scottish socialist Banksy would be spinning in his grave. I speculate this is why the Banks estate pulled out of a deal for Amazon to make a TV version of Consider Phlebas.

[stevebrooks]

Google lied over the capability of its AI!

A video showcasing the capabilities of Google's artificial intelligence (AI) model which seemed too good to be true might just be that.

The Gemini demo, which has 1.6m views on YouTube, shows a remarkable back-and-forth where an AI responds in real time to spoken-word prompts and video.

In the video's description, Google said all was not as it seemed - it had sped up responses for the sake of the demo.

But it has also admitted the AI was not responding to voice or video at all.

In a blog post published at the same time as the demo, Google reveals how the video was actually made.

One wonders why they bothered at all, did they think it was never going to be revealed?

https://www.bbc.com/news/technology-67650807

[Irrev-Black]

It seems that "researcher" should be a job safe from AI replacement, as Charles Stross demonstrates.

The most alarming thing isn't that Bad lies, it's that Bard initially offers up plausible truths, then gradually increases the distortion level as it continues. There's no clear-cut delineation between truth and lies; rather, there's a sliding scale of bullshit.

http://www.antipope.org/charlie/blog-st ... -lies.html

[Irrev-Black]

It seems that "researcher" should be a job safe from AI replacement, as Charles Stross demonstrates.

The most alarming thing isn't that Bad lies, it's that Bard initially offers up plausible truths, then gradually increases the distortion level as it continues. There's no clear-cut delineation between truth and lies; rather, there's a sliding scale of bullshit.

http://www.antipope.org/charlie/blog-st ... -lies.html

And, as is so often the case with Stross bloggery. I'm wasting hours enjoying the comments.

[stylofone]

It seems businesses can pay to have ChatGPT replace help desk staff, but the AI needs a bit of training to stay on task.

Screenshot 2023-12-17 at 12-20-56 Aus.Social.png (157.55 KiB) Viewed 2148 times

There are limits. Writing code is one thing, but we can't have our staff wasting time chatting about Descartes and theology now, can we? That would be unprofessional! The coding though, no problem!

Screenshot 2023-12-17 at 12-21-33 Aus.Social.png (155.58 KiB) Viewed 2148 times

[Irrev-Black]

mettie@froggie.gay
Mettie (Gay Frog Edition)
@mettie@froggie.gay

Presto automation sold an AI chatbot to drive thrus so they could sack their workers while the chatbot would take the food orders. Turns out the chatbot wasn't AI but it was connected to an off shore site in the Philippines where actual humans would pretend to be a chatbot and take the orders.

Arguably, the first case of humans stealing a job from an AI.

https://kolektiva.social/@mettie@froggi ... 6855588849

[Wrenn]

Also, Robots!

[Irrev-Black]

Also, Robots!

Saith Wiki:

Expert opinions

Reactions across the robotics community to Optimus and its prototypes have been "diverse", with many experts commending the venture while describing early demonstrations as less than impressive.[19]

Carl Berry, a lecturer of robotics engineering, described the 2021 AI Day presentation as "the usual overblown hype."[15] Following the Tesla Bot display at the Cyber Rodeo event, researcher Gary Marcus stated he would "bet that no robot will be able to do all human tasks by the end of 2023."[3]

Regarding the second AI Day presentation, Deutsche Welle cited experts - AI researcher Filip Piekniewski, robotics expert Cynthia Yeung and executive director of Mass Robotics Tom Ryden - calling the project a "complete and utter scam", questioning how advanced it really was, and criticizing the choice of a humanoid form.[20]

[stylofone]

The Tesla robot walks as if it's going to the emergency department to have a sex toy removed from one of its orifices. You'd think the designers would have thought of that in advance. If it can't deal with the basic needs of its likely customers, it won't sell very well.