Claude 3 is Anthropic’s new LLM and we believe it’s the best chatbot on the market right now. It’s got three different models but the free version, Sonnet, is excellent and Opus, their paid model, feels like the next step in terms of AI. Of course, all of these models…
@anthropic-ai's Claude 3 has just come out and it is GOOD. So we here at AI For Humans decided to test it the best way we know how, by having it face off in a new AI DEBATE. We used Claude 3 and @OpenAI's GPT-4 to hallucinate two personalities and…
HIT THAT SUBSCRIBE BUTTON MEATBAGS This week… Elon Musk sues @OpenAI, Anthropic’s Claude 3 is the best LLM we’ve used, @Google's Sergey Brin knows they screwed up & AI tax bots gone very wrong. Plus, Gavin dives into text-to-image newness with Ideogram 1.0, Kevin shows off Dust3r which does simple…
Robotic Systems Lab from ETH Zurich has released an incredible new video of their robot, which we affectionately refer to as “Wheel Bot”, which can move fast and toss boxes with aplomb. WheelBot uses a reinforcement learning algorithm to learn as it does things over time and they’re already showing…
SUBSCRIBE TO THE CHANNEL RIGHT NOW This Week In Robots is our new segment where each week (maybe) we cover some of the coolest new robots out in this here AI world. We got updates from @tesla's Optimus Robot, @EngineeredArtsLtd Ameca, @figureai's Figure 01 doing some box sorting & our…
FIRST AND FOREMOST, YOU MUST SUBSCRIBE. This week… @Google Gemini is biased. Now what? Plus, @Stability_AI's Stable Diffusion 3 is here and @NVIDIA's CEO Jensen Huang says that programming as a job might be dead. Then, Kevin made Guy Fieris & Grandmas using GUI for Gligen, Gavin shows off an…
GLIGEN is an innovative way to place images in specific Stable Diffusion generations and now there’s a very simple GUI (Graphical User Interface) to use it. If all of that sounds like complex word spaghetti, the ELI5 version is that this lets you basically create boxes in a two dimensional…
@KevinRose visited the AI For Humans lab and as soon as we turned on PitchBot, we realized what a terrible mistake we'd made. SUBSCRIBE TO OUR CHANNEL Y'ALL Kevin Rose is well known for his investments in companies like Twitter and Uber and now he's faced with an AI For…
So... @Google Gemini is having big issues but here's a weird small hallucination that made us laugh. Google Gemini Pro 1.5 is coming out and people are going crazy for it but our experience with Gemini Ultra 1.0 (c’mon with these names Google) has left something to be desired. We…
OpenAI keeps dropping new Sora videos and this one really had us shocked in that neither of us first thought it was actually AI video. Someone clearly put their dog in a hoodie. RIGHT? No, this was just another example of how AI content can fool us, especially on smaller…
Is it over? Nah, it's just gonna be WAY different. Kevin & Gavin explain what @OpenAI's Sora text-to-video product is, how it works (the dummy edition) and then discuss what happens next. The implications of it being not just text-to-video but also a version of a 'world simulator' with emergent…
This week… Sora is @OpenAI's new text-to-video model and it can do a lot, @reddit is licensing its data for AI & @Google_DeepMind's Gemini got ANOTHER update. AND THEN… an interview with comic book writer Gerry Duggan! He is NOT a fan of generative AI but Kevin and Gavin use…
OpenAI’s SORA model took everyone by surprise and while we’ve yet to get hands-on with the text-to-video AI model, there’s some interesting other videos buried on OpenAI’s website about the video-to-video capabilities of SORA. It’s remarkable to see how it can handle style transfers across the board and, of course,…
Goody2.ai is a hilarious new website that plays with the idea of large language models refusing to answer specific queries by making up reasons that it cannot answer anything. This is a live and interactive site and you can go play with it now but it does raise a ton…
OpenAI just announced Sora, its new text-to-video AI model and we’re not sure the world is ready for this. It feels so far past everything we’ve seen before and while we haven’t been able to try it yes, we’re salivating to see how well the coherence stands up over the…
This week…we dive into @Google Gemini Ultra 1.0, @OpenAI's Sam Altman's making AI chips, @NVIDIA's got chatbots and creepy autonomous AI robots are silently coming for us all.. Kevin explores the new @Stability_AI's Stable Cascade image model, Gavin goes deep on AI music from Suno AI, @elevenlabsio is going to…
An artificial intelligence (AI) model has made a groundbreaking discovery in understanding human language learning by studying headcam recordings from a fraction of a baby's life. The AI, developed by researchers including Wai Keen Vong from New York University, successfully recognized words such as ‘crib’ and ‘ball’. This study highlights…
Media-To-Face is a new AI model that works to better match human facial expressions when speaking or even singing. It’s a very cool new piece of tech that we’re excited to see make it’s way into a ton of other products but… the creepy, gray lifeless face might not be…
@KevinRose stopped by the show to tell us how ChatGPT has a ton of use cases and the things you can do with it keeps growing but it could also save your life. Or at least give you enough basic knowledge so that you don’t really mess yourself up. Being…
This week… @amazon's new shopping AI, a $25m dollar deep fake scam, @Google Gemini Ultra is on the way, Kevin gives us an @Apple Vision Pro review and MUCH MORE. Gavin tells us about a cool new Stable Diffusion plug-in from Glif, @BostonDynamics scary new robot, @HuggingFace's new AI chatbots,…
This week… AI deepfakes cause havoc at X & @Microsoft, @Apple is training a new AI for Siri and it’s using @OpenAI and…. Elon’s gonna put a chip in our brains. YES, FOR REAL. Plus, Gavin tells us about the new AI Arc Search app, Kevin dives into a new…
Act Anywhere is a new AI white paper from the people at Stanford and Adobe which allows you to transfer video into new locations and the video will become subject aware — meaning that if you have someone doing something in one environment and place them in a new one,…
Meta & Mark Zuckerberg have gone DEEP into the AI realm coming out an publicly saying they are currently training their next big model to be open source but that they are truly shooting for AGI. On top of all this Zuck mentioned that the company holds 350k Nvidia H100s…
@RunwayML just launched an update to its new motion brush and it’s really good. Instead of just one brush, you now have five that you can use in a “multi-brush’ array and it lets you create a much more incredible sense of movement as you animate your pictures. As AI…
