11 ms·
Microsoft Bing AI Image Creator
- tluyben2 4y ago> "Image creation is coming soon to your region" Lovely.
- voisin 4y agoWhat would be causing this? Are they licensing something regionally… like Netflix?
- kyriakos 4y agoUnlike Apple, both Microsoft and Google don't offer their services everywhere.
- realce 4y agoExport controls?
- VyseofArcadia 4y agoIf I had to guess (as an ex-MS employee, but not on Bing), legal and compliance.
- BadBadJellyBean 4y agoDoes it work for anyone? Even with a VPN it shows me the same message.
- siraben 4y agoNow that models like Stable Diffusion have been available for months now, allow completely uncensored content and have a rapidly evolving ecosystem around them, are releases like this banking on reaching people unfamiliar with the technology first? Who are the expected users?
- nickthegreek 4y agoRegular people and their kids do not need uncensored content. In fact, an errant boob is the opposite thing these companies want to popup for the general public.
- gjsman-1000 4y agoNot all censorship is an evil.
- lordleft 4y agoIs Microsoft the biggest investor in OpenAI? I wonder what the breakdown is in getting the ability to leverage technology of a company you don't outright own but have heavily supported.
- dariusj18 4y agoThey licensed GPT3 exclusively https://blogs.microsoft.com/blog/2020/09/22/microsoft-teams-up-with-openai-to-exclusively-license-gpt-3-language-model/ https://blogs.microsoft.com/blog/2020/09/22/microsoft-teams-...
- esskay 4y agoI'd assume they are. IIRC they are giving free compute to the project, and those models cost an absolute fortune to train.
- andybak 4y agoThis isn't Stable Diffusion. It's closer to Karlo/Dall-E from my brief tests. EDIT - it's handling some Imagen prompts better than Dall-E. This is interesting.
- andybak 4y agoSo. I was totally wrong. It’s just Dall-e. confirmation bias in action…
- hiddencost 4y agoPragmatically, OpenAI was a mechanism for Microsoft to try risky things with AI without incurring the liability and reputational risk of doing it themselves. They became very conservative after Tay.
- PurpleRamen 4y agoIs this site auto-translated? The first sentence ("Create images from words with AI" in English) is utterly broken and makes barely any sense in my language. Not sure how much I should trust someone who even fails on such parts.
- elaus 4y agoSame for German, totally wrong grammar, so (hopefully?) auto-translated.
- cuuupid 4y agoGiven how many search-paired images Bing has I'd expect them to have a much better dataset and much better quality creations, so this is a little disappointing. So far the best I've seen is B^Discover (uses Kakao's Karlo model) which is a bit surprising because Kakao has far, far, far, far fewer resources in every regard: compute, cash, data, people. (Kakao also open sources almost all of their work.)
- andybak 4y agoWhat prompts have you tried? I'm testing out the very challenging Imagen example prompts: https://imagen.research.google/ https://imagen.research.google/ So far it's doing better than Dall-E/Karlo and a lot better than Stable Diffusion.
- ddmma 4y agoI presume now google users will search for the bing as their top searched keyword
- sourcecodeplz 4y ago"our most searched term is Comet"
- xnx 4y agoHow does one choose between DALL-E, Stable Diffusion, Midjourney, Bing AI Image Creator, etc.? My bias is to go with Stable Diffusion because of its open source nature. Is there an authoritative, up-to-date, comparison of the different capabilities and outputs of these tools?
- andybak 4y agoTry them all. They all have strengths and weaknesses. I use all of them for different things (and often run a prompt across several to decide)
- practice9 4y agoStable Diffusion is the only one you can host yourself or even run locally on desktop & mobile if your hardware allows. You can train your own styles and concepts using Dreambooth method (aka fine-tuning on a very small dataset) for 1.x models, or use textual inversion / LoRA / embeddings for 2.x models. SD2 is much more difficult to fine-tune so folks have been using these alternative methods. Embeddings can represent a specific style and are easier to compose. They are storage-efficient as well (<= 1 MB if I remember correctly). DALL-E is good for surreal memes or some kinds of art. It works well for outpainting. For other use-cases it falls behind on realism and flexibility. Midjourney has come a long way and generates very aesthetically pleasing images. But it's very opinionated to allow for ease of use for new users. And it can get expensive if you need to generate many images.
- Terretta 4y agoFirst line of FAQ says Image Creator is Dall-E. For the outputs, go to each one's community showcase page, or their respective discords, or their subreddits which amount to image post feeds. You'll find, roughly, Dall-E images feel clunky, Stable Diffusion are differentiated, and Midjourney are simultaneously aesthetic and yet recognizably the same. - If you want a sketch of a thing you're thinking of, go with Dall-E. - If you want to mess around, and play with other people's models, Stable Diffusion. - If you want to role play being good at DeviantArt, MidJourney. Quite often, actual artist workflows use two of the three, for instance, start in MidJourney, then outpaint with Dall-E, or start with Dall-E, then img2img with SD.
- freediver 4y agoThe design differnce between bing.com/images and bing.com/images/create is striking. Will AI force the complete redesign of the rest of search UI?
- dannyw 4y agoMeanwhile, Google's answer to this is still locked behind close doors.
- sourcecodeplz 4y agoFirst ChatGTP, now this. Geez, wouldn't wanna be Google right now.
- frozencell 4y agodid google invent the diffusion or this bing does not use it?