SleekEagle

12 ms·

91.
by SleekEagle 4y ago
Upon first inspection, Parti is not as good. This is perhaps unsurprising - in DALL-E 2 the prior model tested between autoregressive and diffusion models and the diffusion model outperformed
92.
by SleekEagle 4y ago
Looks cool although not as good as Imagen. Autoregressive vs Diffusion i guess
93.
by SleekEagle 4y ago
It's not about generating more revenue, it's about cutting costs. Any company that employs graphic designers etc. will be able to cut 90% of the staff. Video game companies that need concept art? How about 1 guy/gal with Imag
94.
by SleekEagle 4y ago
Presumably Google's terms of service or fair use laws. The real restriction is that, even if you had the dataset, training costs tens of thousands of dollars. Only corporations can really afford to train these things. Regarding music -
95.
by SleekEagle 4y ago
I mean DALL-E 2 was the first time my jaw really hit the floor, although in fairness GPT-3 probably should've done that, but it's easier to do with images. And then for this to drop just a month later? Insane. It makes you wonder
96.
by SleekEagle 4y ago
And perhaps even more interestingly these things not only exist but there is competition in this space! Essentially unregulated competition as well (and likely for the next 10 years). The cost will be driven into the ground.
97.
by SleekEagle 4y ago
Do you have a link to any papers about this? Would love to check them out
98.
by SleekEagle 4y ago
It could also be used for more nefarious reasons like disinformation campaigns though... it will be interesting to see what the next few years have in store
99.
by SleekEagle 4y ago
I'm not sure there has been a period of more rapid development in DL than Diffusion Models (maybe transformers?). The next few years will be really interesting.
100.
by SleekEagle 4y ago
I've gotten a lot of "wow, that's cool!"s, which is a pretty fair response for a non-technical person if you ask me!
101.
by SleekEagle 4y ago
How do you think it prints the images!
102.
by SleekEagle 4y ago
For those who use Python but still want to do differentiable programming (and don't want to learn Julia), check out Google's JAX package for Python https://www.assemblyai.com/blog/why-you-should-or-shouldnt-b.
103.
by SleekEagle 4y ago
We demand confidence intervals!!
104.
by SleekEagle 4y ago
I don't think anyone is saying that humanity is close to AGI, but check out DeepMind's Gato work for a more well-rounded agent: https://www.deepmind.com/publications/a-generalist-agent
105.
How Imagen Works(assemblyai.com)
142 points by SleekEagle 4y ago|97 comments
106.
by SleekEagle 4y ago
In this formulation, no, you have no control over the output other than the fact that it is similar to your training data. If you need to have control over the generated image, you would need to use a conditional diffusion model. https:&#x
107.
by SleekEagle 4y ago
Hi there! I'm actually the author of the main article. I actually didn't reference the second article you linked in my writing (I did see it, but did not go through it as I thought it was just a basic summary. In hindsight, I wish
108.
Introduction to Diffusion Models for Machine Learning(assemblyai.com)
98 points by SleekEagle 4y ago|10 comments
109.
by SleekEagle 4y ago
Why are the two necessarily unrelated? Can human being just be considered to be learning via optimization, and perhaps consciousness is an emergent property of an agent with a large enough world model, or a world model that includes the age
110.
by SleekEagle 4y ago
Ultimately it comes down to gradient-based descent (which is pretty magical in its own right), but what's most surprising to me is that the loss landscape is actually organized enough to yield impressive results. Obviously the difficul
111.
Learning to think critically about machine learning(news.mit.edu)
153 points by SleekEagle 4y ago|94 comments
112.
Photorealistic Monocular 3D Reconstruction of Humans Wearing Clothing(phorhum.github.io)
2 points by SleekEagle 4y ago|0 comments
113.
by SleekEagle 5y ago
Thank you for that link!
114.
by SleekEagle 5y ago
I think it's worth pointing out that DALL-E 2 mimics the style of famous artists. The artists had to come up with the original style in the first place! There are highly competent artists that can create highly convincing copies (fab
115.
by SleekEagle 5y ago
You can try to reverse image search - from what I've seen of other people doing this, the renditions are quite distinct. The diffusion process is ultimately the root of the model's ability to not just copy images. Variational meth
116.
by SleekEagle 5y ago
If you have a bit of background in math, I would encourage you to read the CLIP paper: https://arxiv.org/abs/2103.00020 Ultimately, the link between words and their representations comes from the CLIP training. The mod
117.
by SleekEagle 5y ago
As far as I know, OpenAI has not made this dataset publicly available. IIRC the dataset is images scraped from instagram and their corresponding captions. Check out the CLIP paper for more details: https://arxiv.org/abs&#x2F
118.
by SleekEagle 5y ago
The rate of advancement over the past 10-15 years really has been incredible. Now the question is - is this growth curve logistic or exponential!
119.
by SleekEagle 5y ago
Sometimes at the same time!
120.
by SleekEagle 5y ago
Yes! So many exciting possibilities in so many industries. Hopefully it won't displace artists though, we'll have to find a way to manage the efficiency of AI with the curation of artists!

More ›