4 ms·

> I don't think it will improve much from here That's a bold prediction. Why do you think that? The first thing I thought was the exact opposite. This isn't v

> I don't think it will improve much from here That's a bold prediction. Why do you think that? The first thing I thought was the exact opposite. This isn't very good, but it's only version 1. Motion pictures are less than 150 years old. In another 150 years I bet virtual filmmaking will progress a lot.

dgritsko 4y ago
I agree, it feels like this technology is progressing by leaps and bounds almost by the day.
- paulgb 4y ago
  Case in point, just compare the results of Stable Diffusion / Dall-E / midjourney to papers from 5-10 years ago (here's a random one I found http://proceedings.mlr.press/v48/reed16.pdf http://proceedings.mlr.press/v48/reed16.pdf). Remarkable progress in that time, and as I understand it, a good amount (but certainly not all) of the improvement comes “free” from being able to scale up to more weights.
javajosh 4y ago
I don't agree with the prediction but I will say that at least in America, we've been overly optimistic about almost everything since world war II. Is particularly painful when you look at the imagination of science fiction versus the reality of science in the realm of say space travel, especially interstellar travel. This imagination effect gets the public behind your new invention in part because it seems to be just the beginning the start of something new. The tip of the iceberg. But sometimes it's really just a tiny bit of ice. For the record, I'm actually rather bullish on self-driving cars. There's nothing physically impossible about solving the problem, but I'm not surprised it's harder than it sounds. But I don't see humans being fundamentally prevented from solving the problem in the same way that humans are fundamentally prevented from ever engaging in everyday space travel.
mikepurvis 4y ago
Absolutely, particularly if you see this as simply another tool in the belt, like how Jurassic Park's thirty year old special effects absolutely stand the test of time because of them being a convincing mix of early CGI and puppetry. It's not hard to imagine that this kind of thing could end up doing a lot of the heavy lifting for things like background scenes in the future, opening up the kind of stuff we saw in The Mandalorian, Game of Thrones, and the LOTR film trilogy to increasingly lower and lower budget productions.
- vsareto 4y ago
  Most of the AI generated images are good enough to be first drafts or sketches, and with some editing, they can be coherent images. This makes AI generators a good tool. Not sure the same is true for video though. It feels like you'll need reliable still image creation before you can get to a video generator that's useful since editing video takes a bit more than a still image.
- MichaelCollins 4y ago
  The first Jurassic Park holds up unusually well. They used CGI for things they couldn't convincingly do any other way, and used practical effects for almost everything else. And for the most part they hid the CGI well, they took a lot of care to mix it up with close ups of practical effects, the way they framed the shots to direct audience attention, etc. I think in a lot of modern stuff they go too far. They now use CGI for things that could easily be practical effects, but they go with CGI because it's simply cheaper or because they want to A/B test different colors of wall paneling behind a character in post production, or who knows what. The end result is apparently good enough to make a billion dollars, but I can't stand it. Movies don't feel authentic anymore. It's hard to describe rationally, but the word 'soulless' sums up how movies made the new way make me feel. Even the scenes which are wholly practical/real get degraded; the excessive CGI and compositing used in the rest of the movie cast a miasma of unrealness across the entire movie. https://64.media.tumblr.com/248d25e2185a58bf827d329490480fb9/55a1eb08812b2f02-d3/s500x750/c1b6e7b7b6e21956bd7f60f7195794ea1e4998e9.jpg https://64.media.tumblr.com/248d25e2185a58bf827d329490480fb9...
  - 4y ago
mortenjorck 4y ago
The key to predicting technological advancement, that is so often misunderstood by technologists, is the S-curve. Laypeople tend to think everything is linear, technologists tend to think everything is exponential; more often than not, reality tends to be sigmoidal. Technologies have exponential takeoffs followed by logarithmic plateaus. We're clearly well into the exponential phase of deep-learning ML, but it's only a matter of time before this approach hits its logarithmic phase. Of course, the hard part of sigmoidal prediction is determining where in the curve we are. Does the current paradigm have an even steeper part ahead of it? Maybe. And yet, we could just as easily be right in the middle of the function, with a leveling off coming as the state of the art gives way to incremental improvements.

dgritsko 4y ago

I agree, it feels like this technology is progressing by leaps and bounds almost by the day.

paulgb 4y ago
Case in point, just compare the results of Stable Diffusion / Dall-E / midjourney to papers from 5-10 years ago (here's a random one I found http://proceedings.mlr.press/v48/reed16.pdf http://proceedings.mlr.press/v48/reed16.pdf). Remarkable progress in that time, and as I understand it, a good amount (but certainly not all) of the improvement comes “free” from being able to scale up to more weights.

javajosh 4y ago

I don't agree with the prediction but I will say that at least in America, we've been overly optimistic about almost everything since world war II. Is particularly painful when you look at the imagination of science fiction versus the reality of science in the realm of say space travel, especially interstellar travel. This imagination effect gets the public behind your new invention in part because it seems to be just the beginning the start of something new. The tip of the iceberg. But sometimes it's really just a tiny bit of ice. For the record, I'm actually rather bullish on self-driving cars. There's nothing physically impossible about solving the problem, but I'm not surprised it's harder than it sounds. But I don't see humans being fundamentally prevented from solving the problem in the same way that humans are fundamentally prevented from ever engaging in everyday space travel.

mikepurvis 4y ago

Absolutely, particularly if you see this as simply another tool in the belt, like how Jurassic Park's thirty year old special effects absolutely stand the test of time because of them being a convincing mix of early CGI and puppetry. It's not hard to imagine that this kind of thing could end up doing a lot of the heavy lifting for things like background scenes in the future, opening up the kind of stuff we saw in The Mandalorian, Game of Thrones, and the LOTR film trilogy to increasingly lower and lower budget productions.

vsareto 4y ago
Most of the AI generated images are good enough to be first drafts or sketches, and with some editing, they can be coherent images. This makes AI generators a good tool. Not sure the same is true for video though. It feels like you'll need reliable still image creation before you can get to a video generator that's useful since editing video takes a bit more than a still image.
MichaelCollins 4y ago
The first Jurassic Park holds up unusually well. They used CGI for things they couldn't convincingly do any other way, and used practical effects for almost everything else. And for the most part they hid the CGI well, they took a lot of care to mix it up with close ups of practical effects, the way they framed the shots to direct audience attention, etc. I think in a lot of modern stuff they go too far. They now use CGI for things that could easily be practical effects, but they go with CGI because it's simply cheaper or because they want to A/B test different colors of wall paneling behind a character in post production, or who knows what. The end result is apparently good enough to make a billion dollars, but I can't stand it. Movies don't feel authentic anymore. It's hard to describe rationally, but the word 'soulless' sums up how movies made the new way make me feel. Even the scenes which are wholly practical/real get degraded; the excessive CGI and compositing used in the rest of the movie cast a miasma of unrealness across the entire movie. https://64.media.tumblr.com/248d25e2185a58bf827d329490480fb9/55a1eb08812b2f02-d3/s500x750/c1b6e7b7b6e21956bd7f60f7195794ea1e4998e9.jpg https://64.media.tumblr.com/248d25e2185a58bf827d329490480fb9...
- 4y ago

mortenjorck 4y ago

The key to predicting technological advancement, that is so often misunderstood by technologists, is the S-curve. Laypeople tend to think everything is linear, technologists tend to think everything is exponential; more often than not, reality tends to be sigmoidal. Technologies have exponential takeoffs followed by logarithmic plateaus. We're clearly well into the exponential phase of deep-learning ML, but it's only a matter of time before this approach hits its logarithmic phase. Of course, the hard part of sigmoidal prediction is determining where in the curve we are. Does the current paradigm have an even steeper part ahead of it? Maybe. And yet, we could just as easily be right in the middle of the function, with a leveling off coming as the state of the art gives way to incremental improvements.