DPREVIEW closing April 10

He's one of the best with his writing/reviews, I once even told him that (but he is part of newsshooter, or did you mean you don't read CineD?).

Well, you mentioned both newsshooter and CineD. I don't read CineD as much, although I glance at their headlines and read their lab tests on cameras that interest me.

A lot of the value of both those outfits is the testing and hands-on stuff they do. You could write up their conclusions using AI, but I'm not sure how it would speed their review process.

I wanted to post about a paper from openAI and some university about the occupations with highest and lowest exposure to automation by LLMs. Obviously, it should be taken with a grain of salt since it comes from an AI company and also a lot of their methodology is based on guessing.

But the long story short is that jobs that involve critical thinking and jobs that involve physical activity are less vulnerable to LLM automation. Testing cameras (adversarially rather than to market them) involves both critical thinking and physical activity.

(Motion picture and sound recording is listed in the middle of the pack with regard to automation vulnerability. Things like data entry were at the top, and things like wood working were at the bottom.)
 
surprisingly, I don't really follow what's been going on, but I imagine there will be a lot of change on a monthly basis (what people are saying now/are sure of will be completely wrong maybe 6 months later).
 
this recent AI wave and how it's heating up kind of reminds me of the craziness of Covid, i don't know if anyone really knows anything yet
 
surprisingly, I don't really follow what's been going on, but I imagine there will be a lot of change on a monthly basis (what people are saying now/are sure of will be completely wrong maybe 6 months later).

I'm sure that's true, but the basic contours of the technology are taking shape. For example, if generative AI can create text-to-video with low resolution and low framerate, then I expect that they will be able to increase the resolution and framerate in short order. But if LLMs historically and consistently struggle when confronted with novel challenges that are unlike their training data, then I expect that weakness to persist until a more suitable type of AI is invented.
 
this recent AI wave and how it's heating up kind of reminds me of the craziness of Covid, i don't know if anyone really knows anything yet

I'd say that's true. And I'd like to refrain from making concrete predictions. But when I see something like these new AI headshot package services that look usable to me in the here and now, I'll call it. Score one for the machines.

People are also saying LLMs can help them write code in languages they don't know.
 
Not entirely sure what you mean about reverting to originality. But AI generally seems to me like a conformity machine. That's the basis of how it produces seemingly human responses -- by predicting the likeliest output

It means that, in this instance, a professional gear review that can't be copied to a tee by a computer. In other words, insert silly jokes at will because that's where the AI interface will stumble and will stumble for a long time. The "likely output", as you frame it, is a flat, rigid, conventional sentence of a paragraph.

Now, when you're talking about text-to-video, conventional may be exactly what a screenwriter has in mind. He may not want to surprise anyone with his directing ability, just something that makes a play come alive.

PS. Somewhere, long, long ago, i was watching an interview with Larry David about directing CYE. He was asked, "so you mostly do a two-shot"? And he shrugged his shoulders and said, "Ya, a two-shot".


And if an AI, or whatever it's called, can do a two shot text-to-video directing, then it'll create a tidal wave.
 
It means that, in this instance, a professional gear review that can't be copied to a tee by a computer. In other words, insert silly jokes at will because that's where the AI interface will stumble and will stumble for a long time. The "likely output", as you frame it, is a flat, rigid, conventional sentence of a paragraph.

As time goes by, I'm not sure anyone will want to read this stuff anymore. Especially if there's a glut of material.

Now, when you're talking about text-to-video, conventional may be exactly what a screenwriter has in mind. He may not want to surprise anyone with his directing ability, just something that makes a play come alive.

Sounds like you're talking about the low end of cinema, but sure.

PS. Somewhere, long, long ago, i was watching an interview with Larry David about directing CYE. He was asked, "so you mostly do a two-shot"? And he shrugged his shoulders and said, "Ya, a two-shot".


And if an AI, or whatever it's called, can do a two shot text-to-video directing, then it'll create a tidal wave.

The heart and soul of CYE is in the writing, not the directing, acting, editing or cinematography. (There's some good acting.)

Also, I think he's selling himself short. Here's a highlight reel from season 5 where you can see a diversity of shot types and how the cinematography gets creative to deliver visual gags
 
As time goes by, I'm not sure anyone will want to read this stuff anymore. Especially if there's a glut of material.

A glu? . Is there a glut of words in "War and Peace"? And has anyone read it?

Not unless it's a 9th grade assignment for the summer. And even then.

Sounds like you're talking about the low end of cinema, but sure.

It may not even be a cinema per se. It just may be useful for a pitch or just something to entertain yourself.

Murray Rothbard, who's mostly known as an economist, once wrote a satirical one act play "Mozart was a red" about his adventures in the Ayn Rand circle. For a guy who mostly wrote long economics articles and textbooks on very dry subjects like the banking system, the play was surprisingly clever. Now, it can be turned into a YouTube clip. Without actors. Just an AI.

Also, I think he's selling himself short. Here's a highlight reel from season 5 where you can see a diversity of shot types and how the cinematography gets creative to deliver visual gags

He knows that too (he has other directors, like his old pal Larry "The Limo/The Fixup-Bris" Charles contributing). Plus, he's been at it for a long time.
 
I don't read newsshooter because they publish quickly. I read Matt Allard because he's incredibly thorough and brings his professional experience to bear. And he's honest and doesn't hallucinate

I think news shooter got bent into being a stream of copy and paste press releases.


Unlike cine D it announces its mission.. 'shooting news'

Yet fails (for example) to slag off any camera that does not have preroll, which is super useful if you shoot news, critical really.

Its reviewing arri new handunit instead. An arri hand unit has few (but some) place in shooting news.

I think these sites have not helped the manufacturers move on from the dirvel that they try to peddle us.

They note most product to be acceptable which they are not. Well not for shooting news anyway.
 
What a shame. Of course Amazon is going to delete the site and all of that valuable information, knowledge and insight that has been collected over the past 25 years will be gone...
 
I am surprised that the retailers and the manufacturers can't keep a site like DPR up and running? It is basically where people go to make camera and lens purchase decisions. Seems like websites like DPR are more beneficial to have around rather than not. So they cut some staff to reduce costs but to delete it seems odd to me. I would guess that most use the forums a lot more than focus on the content. How expensive is it to do camera & lens reviews and host a forum? Not very much...
 
I wouldn't expect most people to go there, but maybe, IDK.

You also never know what Amazon is thinking...maybe there's a conflict of interest somewhere (or will be in the future). Because they of course could afford to keep it up-and-running even if it's not recommended based on how corps are structured to operate and heed.
 
I wouldn't expect most people to go there, but maybe, IDK.

You also never know what Amazon is thinking...maybe there's a conflict of interest somewhere (or will be in the future). Because they of course could afford to keep it up-and-running even if it's not recommended based on how corps are structured to operate and heed.

Maybe it's odd but I've been hitting DP Review for at least 3 years and I do it daily after dinner, like I hit this site and 3 others all in a row. Just for the forums. Takes me all of 5 minutes these days unfortunately. Too many people now are posting topics with titles that sound like creative postulations instead getting to the point - i.e., a question or an opinion.

I've never used DPR for decision making. The articles read to me like brochures. Everything is rosey. The opinions in the forums are vague, misinformed or fan based from my experience. I will watch some YouTube trials and also more recently go to CameraDecision.com because once I'm honed down to one model I then go through a period of discovery for any details I wasn't aware of that might talk me out of it. That's worked well for me on many interests of mine.
 
I think YouTube is the best place of all time because you have thousands of people communicating information in their own ways.

If one is patient and really interested in learning and he/she watches 20-30-40+ videos, you're going to learn a lot, things you might not even expect to learn.

Dedicated, lengthy and thorough reviews on other sites are great, but most of those end up on YouTube on the proper channels as well.
 
Obviously, YouTube stole most of the DPR's "casual customer" business. A live person explaining various tidbits is more credible than a screen name. Even if that live person is Tony Northrup.

But you got to give Tony/Chelsea, Jared the Fro and others deserved credit for being able to turn clicks and eyeballs into a solid revenue stream with their tutorials and camera guides. That's even better than having a Amazon or B&H affiliate status because, instead of 3%-5% commission fee, the channel creators get to keep full 100%. And that's probably something DPR couldn't do without the Amazon approval, even though that's where the beaucoup bucks are.

PS. I have long thought that a big time streamer like Netflix and its ilk should hire popular YouTubers to steal some business away from Alphabet, which doesn't always deal fairly with its content creators.
 
How expensive is it to do camera & lens reviews and host a forum? Not very much...

That’s what I was thinking. For a company like Amazon, it’s not even the proverbial ‘drop in the bucket’. One would think that there has to be something a little deeper. But then again, the bean counters run these mega-corporations, and if they can save a nickel somewhere, they will.

I worked on a series last year(that shot most of the year) and the rights to have supposedly been bought by Amazon, for this year. Rumor is that production will not begin until later in the year, because of the optics if they were spending money on that, while they are making so many other cuts.
 
Back
Top