sophrocyne

5 ms·

1.
USCO "pre-publication" Report is flawed both procedurally and substantively(chatgptiseatingtheworld.com)
2 points by sophrocyne 1y ago|0 comments
2.
by sophrocyne 1y ago
The USCO report was flawed, biased, and hypocritical. A pre-publication of this sort is also extremely unusual. https://chatgptiseatingtheworld.com/2025/05/12/opinion-why-t...
3.
by sophrocyne 1y ago
Time to vibe code the data into "That supa awesome database over there"
4.
by sophrocyne 2y ago
It is. https://news.artnet.com/art-world/invoke-snags-first-ai-imag...
5.
by sophrocyne 2y ago
For context, I'm Invoke's CEO. I'm not going to argue that they're not complicated tools - Invoke and Comfy are both actively used to push the boundaries of what can be achieved in professional generative media. But I&#x
6.
by sophrocyne 2y ago
Free lunch means it doesn’t require training to achieve the results - it solves for it at run time. You can use this in local generative media tools like Invoke or Comfy, which are both open source and free.
7.
by sophrocyne 2y ago
Forgot to post some context. I'm the CEO at Invoke (i.e., the creator of this piece) -- we're one of the longest running projects for open-source image generation, originating from the initial explosion around Stable Diffusion&#x2
8.
First Copyright for a Single Image Made with AI(cnet.com)
5 points by sophrocyne 2y ago|1 comments
9.
by sophrocyne 2y ago
Hey all. I'm the CEO of Invoke - appreciate everyone who has mentioned us in the thread. To OP -- We work with professional artists regularly, and I'm seeing things pick up as more begin to understand the potential for creative co
10.
by sophrocyne 2y ago
Invoke is model agnostic, and supports Flux, including quantized versions.
11.
Invoke 5.0 – OSS Canvas with Layers and SD/SDXL/Flux Support(github.com)
2 points by sophrocyne 2y ago|0 comments
12.
by sophrocyne 2y ago
The Invoke team released regional guidance using IP Adapter a few months ago, which can use color palettes + style transfer mode, along with text prompts and controlnets. Would take a look at that for some inspiration -- The UI is Apache 2.
13.
The Fine Print of the Stable Diffusion 3 License [video](youtube.com)
3 points by sophrocyne 2y ago|0 comments
14.
Fairness and Fair Use in Generative AI – Law Analysis [pdf](fordhamlawreview.org)
1 points by sophrocyne 2y ago|0 comments
15.
by sophrocyne 3y ago
Some perspectives from someone working in the image space. These tests don't feel practical - That is, they seem intended to collapse the model, not demonstrate "in the wild" performance. The assumption is that all content is
16.
by sophrocyne 3y ago
There is a ton you can do to help SOTA AI remain open. Join the community building the tools - Help with UI/UX, documentation, keeping up with the latest, and evangelizing whatever method the team building it has devised to keep it sus
17.
Dept of Commerce Issues Request for Comment on Open Source AI(ntia.gov)
2 points by sophrocyne 3y ago|0 comments
18.
by sophrocyne 3y ago
Thanks for calling us out - I'm one of the maintainers. Not entirely sure we'll be in the Stable Cascade race quite yet. Since Auto/Comfy aren't really built for businesses, they'll get it incorporated sooner vs lat
19.
by sophrocyne 3y ago
You bet - Thanks for the feedback. Glad you're enjoying Invoke! We do have Docker packages hosted on GH, but I'll be the first to admit that we haven't prioritized ROCm. Contributors who have AMDs are a scant few, but maybe w
20.
by sophrocyne 3y ago
Hey there - I'm a maintainer (and CEO) of Invoke. It's something we're monitoring as well. ROCm has been challenging to work with - we're actively talking to AMD to keep apprised of ways we can mitigate some of the more
21.
by sophrocyne 3y ago
Yes, indeed. Lincoln's still an active maintainer. Energy efficiency is key - Especially with some of these extremely inefficient (wasteful, even) features like real-time canvas. Good luck - Let us know if/how we can help.
22.
by sophrocyne 3y ago
Alex - I run Invoke (one of the popular OSS SD UIs for pros) Thanks for your work - it’s been impactful since the early days of the project. Excited to see where we get to this year.
23.
by sophrocyne 3y ago
Yeah, Invoke's nodes/workflows backend can be hit via the API. That's how the entire front-end UI (and workflow editor/IDE) are built. I'm positive this can be done w/ Comfy too.
24.
by sophrocyne 3y ago
https://github.com/invoke-ai/InvokeAI - runs on Mac silicon, can squeeze out SDXL images on a 16gb mac with SSD-1B or Turbo models.
25.
by sophrocyne 3y ago
There are already a number of local, inference options that are (crucially) open-source, with more robust feature sets. And if the defense here is "but Auto1111 and Comfy don't have as user-friendly a UI", that's also al
26.
Show HN: Invoke AI 3.1 – open-source Workflows and Canvas for SDXL(invoke.ai)
1 points by sophrocyne 3y ago|0 comments
27.
Show HN: Open-Source SDXL UI – Invoke AI 3.0.1(github.com)
1 points by sophrocyne 3y ago|0 comments
28.
Show HN: 3.0 Version of Invoke AI – open-source SD UI and Node-based Back end [video](youtube.com)
6 points by sophrocyne 3y ago|1 comments
29.
by sophrocyne 3y ago
I was able to overcome the simple "word for word" filtering that is being done on book outputs by prompting ChatGPT to write it in pig latin. I succeeded getting the first page of Moby Dick - Chapter 1 (Loomings) - Public domain t
30.
State of the Art: A Perspective from the Frontiers in AI Art(sophrocyne.com)
1 points by sophrocyne 3y ago|0 comments

More ›