AI CLIPPY
Yesterday, Microsoft launched 365 Co-pilot: A suite of AI features that can generate documents and presentations, summarize projects, and more.
It got us thinking: what would an AI Clippy look like? Probably something like this.
It got us thinking: what would an AI Clippy look like? Probably something like this.
💌 Join 500K+ subscribers who get the best of tech every day right to their inbox
PRODUCT HIGHLIGHT
AI Hands are fixed
It’s been a wild week for AI, Google released PaLM, its AI language model, OpenAI stole its thunder by launching GPT-4, Google then released new workspace features, and yesterday Microsoft responded by announcing Co-pilot for 365.
It’s kind of like playing cat and mouse but with world-changing innovations.
Well, take a breath because the week isn’t over.
The photo above? It’s not real. It’s generated with AI by Midjourney.
Midjourney launched last year, and since then, AI art enthusiasts have been filling the model with complex prompts in the hopes of creating even more realistic photos, to the point where Midjourney art has garnered a reputation for being almost too realistic.
It seems the creators of Midjourney took that as a personal challenge when creating the latest iteration of the photo generation engine, V5.
The founders have taken everything they learned over the course of the past year, amplified its strengths, and doubled down on fixing its weaknesses, all in an effort to close the gap between AI and human-generated content.
The latest version comes with higher image quality, more diverse output, seamless textures, support for more natural user prompts, and more. Since its launch, users have noticed more realistic facial and skin textures, cinematic lighting, better shadows, and photo-realistic eyes.
Hands are now fixed: It’s become an online trope that if you’re ever struggling to identify if a photo is AI or not, look at the hands. Time and time again, we see near human-like characters going about their days in a photo only to look at their hands and see a cluster of 7, 10, or even 15 fingers.
It seems like Midjourney fixed this, making it harder for us all to go into detective mode. Needless to say, we may need to be a bit more skeptical about photos we see online. It’s almost as if you can generate…
It’s kind of like playing cat and mouse but with world-changing innovations.
Well, take a breath because the week isn’t over.
The photo above? It’s not real. It’s generated with AI by Midjourney.
Midjourney launched last year, and since then, AI art enthusiasts have been filling the model with complex prompts in the hopes of creating even more realistic photos, to the point where Midjourney art has garnered a reputation for being almost too realistic.
It seems the creators of Midjourney took that as a personal challenge when creating the latest iteration of the photo generation engine, V5.
The founders have taken everything they learned over the course of the past year, amplified its strengths, and doubled down on fixing its weaknesses, all in an effort to close the gap between AI and human-generated content.
The latest version comes with higher image quality, more diverse output, seamless textures, support for more natural user prompts, and more. Since its launch, users have noticed more realistic facial and skin textures, cinematic lighting, better shadows, and photo-realistic eyes.
Hands are now fixed: It’s become an online trope that if you’re ever struggling to identify if a photo is AI or not, look at the hands. Time and time again, we see near human-like characters going about their days in a photo only to look at their hands and see a cluster of 7, 10, or even 15 fingers.
It seems like Midjourney fixed this, making it harder for us all to go into detective mode. Needless to say, we may need to be a bit more skeptical about photos we see online. It’s almost as if you can generate…
A new reality
Yesterday's Top Products
CAT NIPS
- Noizer One is a sound mixer designed for focus, creativity, and tranquillity.
- Animaker Subtitles generates subtitles in 35 different languages through AI.
- UI Quiz lets you put your design skills to the test and see if you’re the next Jony Ive.
- Create variations of images you find online by using CF Spark Browser Extension.
MAKER'S CORNER
- Generate endless product photos with Canvas by Magic Studio.
- DogQ lets makers of all skill levels do cloud testing with no-code.
- Dreamsands lets you extract full code from images or UX screenshots.