Advertisement
SFW just about the kicks!
- Joined
- Sep 16, 2015
- Messages
- 656
- Reaction score
- 2,002
As it pertains to images, I find myself grappling with my feelings. I generally dislike faked photos for getting off, but AI has become so sophisticated that at times there's no way to distinguish between it and real images unless identified as AI. By now I'm sure I've enjoyed content that -- unbeknownst to me -- was actually AI. It's definitely way beyond the days of photo-shopping! All in all, kind of sad that technology has evolved to the point where we accept things that don't exist in reality.
Similarly, I'm skeptical of society's increasing reliance on AI-based information, as it can't always be trusted for accuracy; you really have to check other sources before accepting it as true. I'm old-school and would never use AI for writing -- but that practice is becoming more prevalent and is also potentially problematic (and, in my opinion, just plain lazy). So far, most AI-written material I've seen does include that disclaimer, although not all writers are going to reveal its use.
I agree with Seeker's attitude of treating AI as entertainment and a personal assistant (in a limited capacity, I hope). There is, however, a danger lurking when the majority of people abandon critical thinking and blindly believe everything they see or hear.
This take on AI is my own and from someone who grew up in an era devoid of much of today's technology.
Similarly, I'm skeptical of society's increasing reliance on AI-based information, as it can't always be trusted for accuracy; you really have to check other sources before accepting it as true. I'm old-school and would never use AI for writing -- but that practice is becoming more prevalent and is also potentially problematic (and, in my opinion, just plain lazy). So far, most AI-written material I've seen does include that disclaimer, although not all writers are going to reveal its use.
I agree with Seeker's attitude of treating AI as entertainment and a personal assistant (in a limited capacity, I hope). There is, however, a danger lurking when the majority of people abandon critical thinking and blindly believe everything they see or hear.
This take on AI is my own and from someone who grew up in an era devoid of much of today's technology.
OP
- Joined
- Jun 12, 2003
- Messages
- 8,750
- Reaction score
- 9,987
I'm intrigued with some aspects of AI writing. One of the things I'm working on is image to text. You post one or more images and AI writes a description or story "pictorial" as it were. Some of the output I've read is pretty good. We'll have to see where it goes.
- Joined
- Sep 16, 2015
- Messages
- 656
- Reaction score
- 2,002
I've seen some examples...I would classify this aspect as "entertainment".
- Joined
- Apr 21, 2003
- Messages
- 414
- Reaction score
- 231
AI is a good research assistant, but it needs "Adult Supervision." Sometimes when it does not know the answer, it will make one up. This is known as hallucination. It can easily be tricked into hallucinating, too. It's very gullible and it will believe something if you tell it that it is true.
I coach people on job searching. I recommend that they load the job description into AI and then ask it to provide a list of questions that they may be asked on an interview about this job description. AI does a good job with this. It is also a good tool to research the company to which you are applying. One of the questions I ask is if the company has been involved in litigation. Intellectual property disputes are one thing; EEO complaints and wrongful deaths are another.
As for image generation, the results are mixed. Some are photographic quality in nature. Most are either outright cartoons or enough cartoonish in nature that even a glance shows their AI origin. On the other hand, if you are looking for fantasy illustrations, it can knock it out of the park. (see attached example).
The AI generators I use know what Vans sneakers look like and Crocs. Not so much Keds or Converse. Apparently, there is no way to teach it what these sneakers look like.
I would never use AI for air defense. I've asked it to generate various aircraft images. Even with thousands of images available online, it ALWAYS gets them wrong such as putting props on jet aircraft.
It's a tool. Use it. Trust but verify.
I coach people on job searching. I recommend that they load the job description into AI and then ask it to provide a list of questions that they may be asked on an interview about this job description. AI does a good job with this. It is also a good tool to research the company to which you are applying. One of the questions I ask is if the company has been involved in litigation. Intellectual property disputes are one thing; EEO complaints and wrongful deaths are another.
As for image generation, the results are mixed. Some are photographic quality in nature. Most are either outright cartoons or enough cartoonish in nature that even a glance shows their AI origin. On the other hand, if you are looking for fantasy illustrations, it can knock it out of the park. (see attached example).
The AI generators I use know what Vans sneakers look like and Crocs. Not so much Keds or Converse. Apparently, there is no way to teach it what these sneakers look like.
I would never use AI for air defense. I've asked it to generate various aircraft images. Even with thousands of images available online, it ALWAYS gets them wrong such as putting props on jet aircraft.
It's a tool. Use it. Trust but verify.
Attachments
OP
- Joined
- Jun 12, 2003
- Messages
- 8,750
- Reaction score
- 9,987
I agree but 100 years from now, even 25 with the the strides humans are making. However the thought of Cyberdyne Systems does comes to mind.
Last edited: