advertisement


AI and human communication

tuga

tuga

Legal Alien
I have just watched Grammarly's new advert where it shows the programme virtually writing an important email for the person who should be sending it to her co-workers.
Could the use of AI lead to humans losing their ability to communicate?
 
I see more people using it for sending emails in work situations. Once you've used ChatGPT a few times you can generally spot when people have copied it word for word.
As for losing the ability to communicate - yeah maybe. but that might have started before AI became prevalent. You see people just staring at their screen all the time when really they shouldn't be. I see people out for dinner and they're all like zombies. people can't even put them down to drive. I see so many people flicking their eyes down when i am out walking the dog. I'm guilty of changing since smartphones came out. I used to read so much fiction. I loved it. I read more than anyone i know. haven't read a book in years now.
 
  • Like
Reactions: gez
this has nothing to do with AI - but has everything to do with how intrusive technology in our lives and how we interact with that technology.

My handwriting has become appallingly bad as i stopped having the need to keep written notes for work based things.
 
I’m looking forward to the day that work emails are written by AI as long as those emails are also read by and replied to by AI. My hope is that the technology advances to the point that AI spends the day “reaching out” to other AI and adding each other to cc fields so I can be left alone to work (or to watch funny cat videos)* instead of reading and replying to emails.

Joe

* I’ve not yet decided how I’d spend the temporal windfall.
 
gintonic said:
this has nothing to do with AI - but has everything to do with how intrusive technology in our lives and how we interact with that technology.

My handwriting has become appallingly bad as i stopped having the need to keep written notes for work based things.
Click to expand...
Same here.
I write like a geriatric now.
And in case you read the hip thread...
Leave it!
 
tuga said:
I have just watched Grammarly's new advert where it shows the programme virtually writing an important email for the person who should be sending it to her co-workers.
Could the use of AI lead to humans losing their ability to communicate?
Click to expand...
I'd argue that humans have already lost their ability to communicate, at least in person. There are generations now who've effectively grown up pretty much only communicating using texting or similar online apps. Their ablity to communicate in person or be anything but extremely defensive and see conflict in every statement has been substantially reduced. I genuinely fear for the world.
 
tuga said:
I have just watched Grammarly's new advert where it shows the programme virtually writing an important email for the person who should be sending it to her co-workers.
Could the use of AI lead to humans losing their ability to communicate?
Click to expand...
As long as there's a graph involved you'll be just fine.
 
Tony L said:
Another thread to drop physicist Angela Collier’s AI video into:


She nails it IMHO.
Click to expand...
I've only watched the first 10 minutes so far. But agree with everything she's said to that point. Will watch the rest later, but it did make me think one thing though:

Artificial Inteligence is actually the perfect name for it. Because it's not real, it's artificial. As in it's not really inteligence at all. Just like artificial plastic plants aren't real plants. So ironically, for totally the opposite reasons to why it was named AI, AI is in fact the perfect name for the technology ;) :D
 
I fear that the world is rapidly heading toward a future that is akin to 'Gilligan's Island' sans the professor.
 
Whether we call it artificial intelligence or machine intelligence on some narrowly defined tasks clever 'puters are already better than humans, even highly trained ones. A student at work was able to show that AI was better than trained radiologists at diagnosing pneumonia from lung ultrasound data.

From https://lfpress.com/news/local-news...-in-lung-ultrasounds-bests-trained-physicians

“We were a little shocked. It had a perfect record of separating the COVID from the non-COVID pneumonia (ultrasounds),” said Robert Arntfield, a Lawson researcher and medical director of London Health Sciences Centre’s critical care trauma centre.​
​
“I’m willing to acknowledge that 100 per cent seems too good to be true, but in the science we’ve done and the way we’ve conducted it, it’s sufficiently strong that we’re very confident there’s a signal there, kind of like a QR code buried in the lungs that we can’t see but the computer vision can extract.”​

The doctor on the study was just floored.

Joe
 
gez said:
'd argue that humans have already lost their ability to communicate, at least in person
Click to expand...

that depends on how much you get out and about, and interact with people...... i would argue that we are far from losing our ability to communicate in person. Communications are evolving, but given my 5000 students - most of whom are young people, the vast majority understand the contexts within which they communicate and evolve to meet the needs of those contexts.
 
Colin L said:
As long as there's a graph involved you'll be just fine.
Click to expand...
It's my day job (communicating with graphs, but also verbally). I think that my command of the English language has gotten pretty good these past few years.
 
Joe P said:
Whether we call it artificial intelligence or machine intelligence on some narrowly defined tasks clever 'puters are already better than humans, even highly trained ones. A student at work was able to show that AI was better than trained radiologists at diagnosing pneumonia from lung ultrasound data.

From https://lfpress.com/news/local-news...-in-lung-ultrasounds-bests-trained-physicians

“We were a little shocked. It had a perfect record of separating the COVID from the non-COVID pneumonia (ultrasounds),” said Robert Arntfield, a Lawson researcher and medical director of London Health Sciences Centre’s critical care trauma centre.​
​
“I’m willing to acknowledge that 100 per cent seems too good to be true, but in the science we’ve done and the way we’ve conducted it, it’s sufficiently strong that we’re very confident there’s a signal there, kind of like a QR code buried in the lungs that we can’t see but the computer vision can extract.”​

The doctor on the study was just floored.

Joe
Click to expand...
Did you hear the part in the YouTube video Tony linked to, upthread, where she talks about AI doing TB diagnostics?

TL;DR:

They taught a ML bot to identify TB from scans using lots of scans. Then they checked the learning using another dataset and the ML diagnosed TB more accurately than a hospital consultant specialist in TB.

But.

ML tools are a black box, you don't know how it taught itself what it did, so you don't know how it concluded what it did. In that particular instance, they managed to get under the hood of the ML tool and found out what it had learned to make its diagnosis. Turns out, it was using the age of the machine that did the scan. Turns out that the older the machine, the more likely the scan was to contain TB. Turns out, that's because the older machines are normally found in poorer countries, where TB is more prevalent.

Beware of trusting ML bots unless you are sure you know what they learned.
 
Steve,

No, I didn't watch the video, but I do know about such research. As a diagnostic tool, AI has incredible potential.

AI is also folding and unfolding proteins, which sounds easy but is anything but. This is a huge breakthrough and will be the genesis of a medical revolution ... for those who can afford it.

Joe
 
You must log in or register to reply here.


advertisement


Back
Top