Just posting a promotional video is not a contribution to open source photography in my eyes, which is what pixls.us is all about.
2 Likes
Fair enough. I just added the link to the site cos it was mentioned in the vid. I can remove if u prefer
This is the discussion section. Any input from your side beside the link to a rather long video?
1 Like
I tried to find a transcript to excerpt some key points that seemed interesting but couldn’t immediately find one. I don’t like what these guys do, particularly, but it does seem relevant to what people do here.
In short they talk about integrating for instance modules to, say, LR that would allow users to “add a necklace to that portrait”, change it in such and such a way and, in a way that’s not clear exactly, go beyond using text prompts to do those kind of transformations. I don’t want to do that myself but I can imagine it would change the nature of all imaging
Isn’t that already in processing applications like Luminar that has AI sky repalcement and such things?
We will see. I personally do not want to create “images” but to makes photographs. For some time it is possible to easily replace the sky in a photograph. However, I never really had urge to do that. At least not seriously.
Certainly, but even factoring in these guys’ VC spin, this stuff might become shockingly good, shockingly fast. Maybe it will be good. A lot of generic manicured landscape photography is already a little bit unbelievable and will seem more so if you can generate it in software.
I’m totally with you, but on the flipside, it took journalists a while to realise that the internet came not to praise them but to bury them.
I think we are already there.
In the long run, I think it will increase demand for human made art things.
Sort of like you can get a Budweiser anywhere, but the craft brew has never been stronger.
6 Likes
That’s possible. Maybe we can already see it in the trend for snapshot type pics. The worse the better, to connote that something is “real” rather than manipulated, though the software can also generate that, I suppose. I just think that once you start messing with business models and economic incentives, you end up in unexpected places, for good or ill. From direct experience, journalists thought the web would expand the market for writing and for a while it did. And then it destroyed 75% of the industry, the most important bit in fact, the local and regional news.
Ugh.
[Sorry, couldn’t let that slide by]
5 Likes
I flicked through the OP 80-minute video of talking heads. If that’s the best that human intelligence can produce, then give me AI every day.
5 Likes
LOL
It’s very pervasive in so many area’s. I read that the Chinese were using it in dogfight simulations against F-35 and it would lose more often than win until it didn’t … Deep fake simulations, self driving transportation etc etc…
At the university plagiarism was already a big problem in the digital era and now it’s become a wholesale scramble to deal with AI generated material being submitted as original work.
I think we are going to see it everywhere… Like most technology, its not the technology itself but more so the speed of the changes that are inevitable and the inability of society, policies, laws, and regulations etc to keep up…
2 Likes
Perhaps I maligned the OP video. Perhaps it was made by AI, with the prompt, "“Make a boring video with two boring blokes saying how fascinating and wonderful AI is. Make the video really boring. Do not include any cutaways with actual examples of AI. Just have the blokes talk boringly about it.”
Some day soon, camera viewfinders will have a dancing paperclip in the bottom-right corner saying, “It looks as though you are trying to make a portrait. Can I help you with that?”
Given that AI does mash-ups of what is on the internet, and AI generates this stuff quickly, the internet will soon have more AI than human content. And eventually all AI content will be mash-ups of AI content. Then the content will be ultra boring.
I reckon human content is (largely) going backwards. I did some TV 45 years ago. It wasn’t great quality – black and white, crash edits by copying from one massive tape deck to another. We would have killed for current technology. But I don’t think any of our product was as boring as the OP video. If AI is as important and fascinating as they think it is, why couldn’t they make an important and fascinating video?
And that brings me to quality. There is some good stuff on the net. But most videos that I see are dull, dull, dull. We are still at the “gee-wizz” stage of video technology, where content-makers think that pointing a camera at something automatically makes an interesting movie. It doesn’t.
AI is certainly still in that phase. “Gosh, this AI image looks photographic.” It isn’t a photograph, just a mashup of other photographs. Perhaps we can see that, perhaps we can’t. But is the AI image interesting? Is it interesting only because it is AI? Are we so dazzled that we don’t ask deeper questions about quality?
3 Likes
Let me get my button presser.
2 Likes
It’s put a damper on my retouch playing for sure; not sure if it is doing the same for photographers. 
You made my day 
!
1 Like