tech

As chatbot sophistication grows, AI debate intensifies

11 Comments
By Julie Jammot and Laurent Barthelemy

The requested article has expired, and is no longer available. Any related articles, and user comments are shown below.

© 2022 AFP

©2023 GPlusMedia Inc.

11 Comments
Login to comment

In October, Character.ai, a start-up founded by former Google engineers, put an experimental chatbot online that can adopt any personality.

It is well nigh impossible to tell if you are chatting with GPT3.5 or a human online. Turing Test passed?

This forum may be populated by chatbot responses.

5 ( +6 / -1 )

Dagon, I think you overestimate the level of the discussion on this forum.

7 ( +7 / -0 )

@timeon

Dagon, I think you overestimate the level of the discussion on this forum.

Well thanks and sometimes I think so. Thanks for noticing anyway , I am just a foreigner living in Japan and you are most probably not a chatbot.

2 ( +3 / -1 )

What the hell is the image?? Looks like some tortured teddy bear!

3 ( +4 / -1 )

Chatbots currently in use for 'customer service' helplines are rubbish. They are easy to spot as they have clear, measured accents, talk over you, speak in whole sentences, deliver copy and paste answers of limited value and are generally unresponsive. The tone doesn't vary and lacks emotion. They are inflexible, unhelpful and impersonal. They are not a threat, just hopeless. Use of a chatbot for customer services is a big red flag indicating that you should avoid the company if you can. I doubt this one will be much better if you have a problem with your bank/tax form/washing machine/health.

4 ( +4 / -0 )

What the hell is the image?? Looks like some tortured teddy bear!

AI art?

I wouldn't trust AI because there isn't an easy way to check that it's telling the trutj or being secretly selective in its presentation of facts (journalists are bad enough).

1 ( +2 / -1 )

I usually do not take sides but one thing is clear AI and Robctics needs to be regulated to the extreme or they will become a Cancer to humanity in the future. This technology is a technology of choice, we take this technology with out regulation and rules of lts use not clearly defined it will become a Cancer that humanity might not have cure for! We highly regulate it and study it for decades and then slowely implement it with out a materialistic thought in mind. We can say no and for go this technology and outlaw it!

Im am for the later, truely against this technology in every way shape or form. It is not my choice to makew

0 ( +1 / -1 )

Yes, AI is a very interesting field, no doubt, but It’s also very much overestimated. I’m deep enough in AI to make that statement under theoretical considerations and having seen much practical evidence. Using AI for boring or repeating and time consuming routine jobs is very fine, some chat responses in customer service, finding X-ray photos with possible cancer, writing some basic computer programming code, mixing many chemical substances for finding a new and stable one, or spitting out very fast what you could also find more slowly yourself on Wikipedia etc, but everything above that level, using and relying on AI on higher sophisticated fields, that will all turn out only as an illusion because AI is capable of nothing then. AI used above basic levels will then only produce itself the data outliers and more severe, misleading or even intentionally manipulated outputs, based on human input data that is not anymore mathematically neutral but very much biased, political more correct, or omitting data as input that some minorities could make feel offended and so on. You liked your playgrounds as a child and found it very interesting and were attracted? Well, AI is such a playground too, but now for very educated adult people, not more, not less, just a fascinating playground.

-1 ( +0 / -1 )

AI already does a good job writing articles on command. They say it will replace content creators, and to a degree this is true. It will replace the lowest level of content creators, writers who engage in content marketing and such, who scrape the internet for info and then regurgitate it in modified form as original articles. This sort of content blankets the internet and is why so many articles sound the same. They are grammatically correct and all, but just lifeless and boring, as well as devoid of original or interesting ideas. AI can do this better than people now.

But it is a long ways from saying something insightful or interesting, so writers who have a voice will be safe. English teachers who assign essays may be facing some trouble though.

0 ( +1 / -1 )

Is the image shown imply that even a dog/ or other animal, can gain intelligence . . . uncanny . . . ?

-1 ( +0 / -1 )

 It will replace the lowest level of content creators

This would be a benefit, trash articles would still be flowing around, but at least the people that currently write them would have to find a more productive way to get money.

English teachers who assign essays may be facing some trouble though.

Probably some are already reading things made this way, it will have to be compensated somehow, for example with tests where the student explains his essays.

0 ( +0 / -0 )

Login to leave a comment

Facebook users

Use your Facebook account to login or register with JapanToday. By doing so, you will also receive an email inviting you to receive our news alerts.

Facebook Connect

Login with your JapanToday account

User registration

Articles, Offers & Useful Resources

A mix of what's trending on our other sites