ChatGPT

Botch

I'm not sure which frightens me more, ChatGPT or deepfake video.  
It started with a clip of Nancy Pelosi sounding "drunk", via slowed down audio track (source: an unemployed MAGAt in his Mommy's basement) then to a video clip of CNN's Robert Acosta "striking" a (faux?) reporter, via sped-up video (source: the cheeto whitehouse).  
I've seen much more finely-tuned deepfake video since, on 60 Minutes and other sources....  if the bottom-third of our population can be bamboozled/swayed by a single loudmouthed shyster who only repeats lies over and over, what happens when video "proof" of QAnon "theories" begin flooding the web?  
Damn.  

JohnInCarolina

We aren’t too far away from universal, near instantaneous translation that’s quite good.

Imagine you’re on a Zoom call with someone in Japan.  They speak Japanese on their side of the call, and your voice and speech is translated from English to Japanese.  Then on your side it sounds like the Japanese person might as well be speaking English on the other side of the world.  

This is Star Trek stuff.   But it’s coming.  

HeavyG

I'd rather have a babel fish in my ear - wouldn't need a computer for translations.

JohnInCarolina

HeavyG said:

I'd rather have a babel fish in my ear - wouldn't need a computer for translations.

Probably not much further away.  But it will still be a computer

dbCooper

Apparently it's not up to doing rocket science yet...

https://www.npr.org/2023/02/02/1152481564/we-asked-the-new-ai-to-do-some-simple-rocket-science-it-crashed-and-burned

lousubcap

I don't buy "simple rocket science equations."  But it will get there, no doubt.  

paqman

JohnInCarolina said:

We aren’t too far away from universal, near instantaneous translation that’s quite good.

Imagine you’re on a Zoom call with someone in Japan.  They speak Japanese on their side of the call, and your voice and speech is translated from English to Japanese.  Then on your side it sounds like the Japanese person might as well be speaking English on the other side of the world.  

This is Star Trek stuff.   But it’s coming.  

We pretty much have that now in Teams live transcripts and live translations.  It is far from perfect but on meetings where transcripts are enabled, participants have the option to translate CC from english to french and vice versa.

nolaegghead

paqman said:

JohnInCarolina said:

We aren’t too far away from universal, near instantaneous translation that’s quite good.

Imagine you’re on a Zoom call with someone in Japan.  They speak Japanese on their side of the call, and your voice and speech is translated from English to Japanese.  Then on your side it sounds like the Japanese person might as well be speaking English on the other side of the world.  

This is Star Trek stuff.   But it’s coming.  

We pretty much have that now in Teams live transcripts and live translations.  It is far from perfect but on meetings where transcripts are enabled, participants have the option to translate CC from english to french and vice versa.

Yep, we use that too.  

Siri, Alexa increase comprehension from machine learning.  The security chip in Apple products, I believe, does machine learning from you.

lousubcap

An interesting commentary regarding ChatGPT:
https://www.newyorker.com/tech/annals-of-technology/chatgpt-is-a-blurry-jpeg-of-the-web?utm

JohnInCarolina

lousubcap

I like the approach. 

However, on the side I figure you are too old for university friends...But in this day and age who knows...

Botch

My favorite type of comedian is the one that not only makes you laugh, but makes you think about things in another way; in that sense, John Oliver is at the top of my list. This is very hilarious, but also very thought-provoking. Enjoy.
 
https://youtu.be/Sqa8Zo2XWc4

nolaegghead

I love John Oliver.  Last Week Tonight is a great show, can't win enough awards.

Botch

ABC News  had a feature story about GPT-4 tonight, which supposedly dwarfs ChatGPT in abilities, just a few months later (same company).  It could pass the Bar exam in the upper 10% (I know nothing about that test or how it's structured), identify not just subjects (my iPhone does that) but other aspects of photographs, diagnose medical conditions (that one's huge), and even use a cellphone photo of your frig's contents and write you out possible recipes (!).  
The two most exciting examples, to me:  providing personalized Individual Learning programs for students.  My ex-GF developed a Nursing Assistance program at the local Community College (which had an interesting mix of brilliant kids and discarded "first wives" who'd never worked outside the home before).  Her program let students work at their own speed, and soon gained national attention; when we were last together she had four  best-selling texts on the market, and did quite well for herself.   
The second: GPT-4 could do your taxes almost instantaneously!  Makes sense, and there's another career field/business this AI stuff will eliminate.  
 
I've always read articles about "guaranteed minimum incomes", over the past couple years, a fascinating possible path for our future.  It would guarantee a minimum payment, for all citizens, to provide for shelter, sustenance, and health care for everyone (there have been a couple cities who've tried this, and it seemed to work well for them, although I didn't save any of those links; it really reduced crime which brought its own benefits).  I know the oligarchs worldwide will scream bloody murder about such an idea, but both BotFC and (just last week) Bill Maher have pointed out that we're slowly but surely moving in that direction already; the change is happening faster than the base culture can absorb.
 
Mechanization, CNC and robotics have greatly taken over "blue-collar" jobs since the 1950s; computers, the 'net and now AI are in the process of doing the same for "white-collar" jobs.  There are some types of jobs that (so far) are resilient (nursing, child/elderly care, early teaching, plumbers) and some of those fields have had jumps in income, some have not (about ten years ago I had a guy rebuild my sprinkling system over a weekend; he made as much money that weekend as I did the previous two weeks, engineering on the space shuttle boosters; but, early/primary teachers are leaving their jobs in droves (no doubt to the delight of the oligarchs)).  
 
There was an ancient Star Trek  episode where Cap'n Kirk said, "We don't have to labor anymore, we can dedicate our lives to Learning/Exploring" or something like that.  There is a wisp of that going on already; look at all the "content creators" on the web, making a really good living helping others Learn/Explore; not something that was done 150 years ago.  We live in Interesting Times.  

Botch

Oh, and not twenty minutes later, I see this:
 

 
  

JohnInCarolina

@Botch - I know a reasonable amount about how these things work, and there is no universe where I’d let one do my taxes.

Botch

JohnInCarolina said:

@Botch - I know a reasonable amount about how these things work, and there is no universe where I’d let one do my taxes.

Well, I assume you're still getting residuals from your pr0n career, so...   

JohnInCarolina

Botch said:

JohnInCarolina said:

@Botch - I know a reasonable amount about how these things work, and there is no universe where I’d let one do my taxes.

Well, I assume you're still getting residuals from your pr0n career, so...   

Good point!

Botch

On tonight's news was a story about an arrest of a guy who planned on executing cops during tomorrow's St. Paddy parade in Yonkers, NY.  Authorities were tracking this guy's posts on social media, and got him before anything bad happened.
There's gotta be a way for AI to be programmed to track every post, every click, every purchase, and every website on the 'net to identify these folks who have been "plotting", or even in the process of being brainwashed, and flag these lost people to local authorities, present the list, and let those authorities make a decision.  
Because of the thoroughness/speed of AI, it could be a much more "secure" way of tracking potential problems, before they act.  We've been bombarded with horror scenarios of how authoritarian types can mis-use AI, there might be some ways to use it for good.  

HeavyG

Botch said:

On tonight's news was a story about an arrest of a guy who planned on executing cops during tomorrow's St. Paddy parade in Yonkers, NY.  Authorities were tracking this guy's posts on social media, and got him before anything bad happened.
There's gotta be a way for AI to be programmed to track every post, every click, every purchase, and every website on the 'net to identify these folks who have been "plotting", or even in the process of being brainwashed, and flag these lost people to local authorities, present the list, and let those authorities make a decision.  
Because of the thoroughness/speed of AI, it could be a much more "secure" way of tracking potential problems, before they act.  We've been bombarded with horror scenarios of how authoritarian types can mis-use AI, there might be some ways to use it for good.  

So basically you want us to become like China where the government is monitoring everything its citizens say and do????

poster

Minority Report was set in the 2050's, not too far off from what your saying 

nolaegghead

Bosch is prescient insofar as we don’t officially have that yet.  China does.

nolaegghead

The bar is different, in either case…for now.  In China you go to a reeducation camp if you’re lucky for criticizing Xi.  Here you would have to be planning a felony.  Ok, so the same.

JohnInCarolina

nolaegghead said:

Bosch is prescient insofar as we don’t officially have that yet.  China does.

Unofficially, the NSA says hello.

dmchicago

Well, thanks @Botch for exposing @JohnInCarolina master plan for all to see. 

Botch

HeavyG said:

Botch said:

On tonight's news was a story about an arrest of a guy who planned on executing cops during tomorrow's St. Paddy parade in Yonkers, NY.  Authorities were tracking this guy's posts on social media, and got him before anything bad happened.
There's gotta be a way for AI to be programmed to track every post, every click, every purchase, and every website on the 'net to identify these folks who have been "plotting", or even in the process of being brainwashed, and flag these lost people to local authorities, present the list, and let those authorities make a decision.  
Because of the thoroughness/speed of AI, it could be a much more "secure" way of tracking potential problems, before they act.  We've been bombarded with horror scenarios of how authoritarian types can mis-use AI, there might be some ways to use it for good.  

So basically you want us to become like China where the government is monitoring everything its citizens say and do????

Dude, we're already there.  Right now its amazoid/googlez that are tracking everything, but they fund/control our congress so it's essentially the same thing.

lousubcap

Some recent concerns regarding AI:

"More than a thousand signatories from Steve Wozniak to Yuval Noah Harari have penned an open letter calling on all AI labs to immediately pause for at least 6 months the training of AI systems more powerful than GPT-4. The experts aren't calling to deep six AI or to send it six feet under, but merely to temporarily eighty-six advancements to give humanity six degrees of separation from a technology that, according to their sixth sense, could change our lives six ways from Sunday. There were plenty of pauses and concerns in the past, but once ChatGPT and Microsoft went public with their offering and got the attention of Joe Six Pack, the race for dollars was underway. Thus, it seems unlikely key players will hit the pause button. Tech companies within arms reach of cash are like sugar-infused kids at Six Flags, and whether anyone waits a few months or not might be a case of six of one, half a dozen of the other. 

Don't feel like reading the whole letter? I asked ChatGPT to summarize it: The development of AI systems with human-competitive intelligence poses significant risks to society and humanity. It is essential to pause the training of AI systems more powerful than GPT-4, develop shared safety protocols, and AI governance systems including regulatory authorities, oversight, and tracking of AI systems, auditing and certification ecosystems, liability for AI-caused harm, and public funding for technical AI safety research. 

I also asked ChatGPT to offer a rebuttal, and it seemed to offer more of a corroboration: The concerns raised in the argument about the risks of AI systems with human-competitive intelligence are shared by many experts in the field. It is important to carefully consider and address the potential risks and challenges associated with the development of advanced AI systems to ensure their positive impact on society. The proposals to pause the training of AI systems, develop shared safety protocols, and implement AI governance systems are important steps towards achieving this goal. "

Kayak

Isaac Asimov might have had a suggestion.

kl8ton

https://finance.yahoo.com/news/joe-rogan-just-issued-warning-104000837.html

JohnInCarolina

kl8ton said:

https://finance.yahoo.com/news/joe-rogan-just-issued-warning-104000837.html

Rogan discovering that his podcast can be replaced by a brainless automaton is just… chef’s kiss.