Videos

Doom can now be entirely generated, in real time, by an AI model. Using Stable Diffusion and a bunch of AI Agents, this group of engineers from Google was able to open the door to an entirely different way of playing video games where the game *could be* generated in…

Join our new Discord! https://discord.gg/muD2TYgC8f AI News: OpenAI’s rumored Strawberry reasoning model is real and might just be paving the way for GPT-5 aka Project Orion according to a huge story in The Information. Plus, GameNGen uses AI to generate DOOM on the fly, Cursor AI is turning everyone into…

Ideogram 2.0 is very good, especially at text but new features like color matching and updated realism give this AI image model a big leg up. Of course, it still has some of the same problems we’ve seen with IP issues.

AI tools are getting more sophisticated but, as always, the simplest way to see if something actually is AI involves using your common sense. Send this video to your parents (or really anyone who needs the basics) around getting educated with what actually is AI media and how to spot…

Hedra and Flux and other AI tools allow you to do all sorts of interesting things but also… allow you to do just the DUMBEST things, like have Baby Joe Rogan interview Baby Terrance Howard. More examples in the longer video attached (around the 28 min mark)

Join our Patreon: https://www.patreon.com/AIForHumansShow AI news coming in HOT: Ideogram 2.0 dropped and, along with Flux community updates & Google’s IMAGEN-3, continues to show how AI imaging tools keep improving… but also opens the door to whole new messes. Plus, OpenAI’s drops new blog posts (wee!), Unitree sends their cheap…

We uncovered an old trove of dating videotapes and whoo doggy were they weird. Created with Midjourney, @RunwayML GEN-3 Turbo, Hedra's Lip Sync, @elevenlabsio's voices, @SunoAimusic1's music and a whole lot of human stupidity. Join our Patreon: https://www.patreon.com/AIForHumansShow #ai #aitools #aivideo // CHAPTERS // 00:00 Meet Chris 01:27 Meet Cynthia…

Grok 2 is actually pretty good and one of the things that makes it great is the integration of Flux, a new cutting edge image model. However, there are some people having issues with the sorts of images it can make and we had a similar experience.

AI NEWS: Xai surprised us all with the drop of Grok-2 with Flux and it’s actually very good, @Google's new Pixel event was AI heavy as @Google_DeepMind's Gemini Voice looks to steal OpenAI’s thunder.@OpenAI *did* drop a new model but it wasn’t the rumored Strawberry…at least not yet. Plus, AI…

The Flux Realism Lora continues to surprise me across all uses of it - and today i got it to report on the news that yes, I totally suck with the help of Hedra. FULL TUTORIAL IN THE LINK

AI Tools: We combine the Flux Realism LoRa with Hedra's super fast AI lip-sync tool and were pretty shocked at how quickly & simply we could make something interesting. Can Old People Drive? I guess they can. PS, this video is not sponsored but we are always listening. Join our…

Flux’s new image model opens the door to all sorts of weird new things — and when you combine them with what’s possible with AI lip-sync now via something like Hedra it gets weirder, faster.

Join our Patreon: https://www.patreon.com/AIForHumansShow AI NEWS: OpenAI is losing co-founders & getting sued by Elon but might be on the verge of a big new model. Strawberry Q* anyone? Plus, we dig in on Figure Robotics new 02 robot, the amazing Flux image (and soon to be) video AI model,…

Flux Pro is a brand new AI imaging model that is both free to try, able to be run locally & open sourced. We put it up against the newest Midjourney 6.1 update to see which one could do better hands, the ultimate test of AI.

AI News: Midjourney 6.1 and the brand new Flux Pro just dropped and Gavin puts them through significant tests around realism, text generation and, of course, HANDS. Oh, an in the meantime, generated some pretty terrible images as well. Join our Patreon: https://www.patreon.com/AIForHumansShow Follow us for more on X @AIForHumansShow…

Flux is a new AI image model from the people who made Stable Diffusion and WOW it’s very, very good especially because it’s open source and free to play with (at least for now). Black Forest Labs, the team behind Flux, worked on the Stable Diffusion models and it’s not…

Meta’a Llama 3.1 is pretty amazing but then when you combine it with the speeds of Groq’s custom silicon you get a chance to see just how insane the future will be. Also, let's rank sausages!

Join our new Patreon: https://www.patreon.com/AIForHumansShow AI News: @meta AI's new LLAMA 3.1 405b parameter model is here, it’s great and Mark Zuckerberg may now be leading the open source AI movement. Not to be outdone, @OpenAI's GPT-4o Mini arrives! Plus a new study on Universal Basic Income & Adobe releases…

Join our Patreon: https://www.patreon.com/AIForHumansShow AI News coming in hot: @Proof_news shows that YouTube transcripts were used to train The Pile from Eluther.ai, an AI dataset that NVIDIA, Amazon, Apple & Anthropic all use, @a16z & other Silicon Valley leaders choose a side in the election, @OpenAI's new model is leaking…

It's an all-out AI safety debate as @OpenAI's GPT-4o & @anthropic-ai Claude imagine their own debate characters and then face-off in an epic and conflicted conversation to decide whether... AI GOOD OR AI BAD?!? Yes, we understand this prob isn't how AI ethics should get decided but you know, it's…

Is generative AI losing steam? We kind of doubt it but... Goldman Sachs of all places dropped a long article called “Gen AI: Too much spend, too little benefit” where they argue that the tens of billions of dollars being fed into Generative AI has not been worth it so…

This new cell phone factory is a “dark factory” meaning because it has no human workers they can operate it without lights on. The robots will make one cellphone a minute entirely on their own which is still below human factories but shows a future where there will be a…

Join our new Patreon: https://www.patreon.com/AIForHumansShow AI News coming in hot: Goldman Sachs issued a report that AI is *extremely overhyped*, OpenAI blocks access to its API in China and the rumors are swirling, Anthropic drops the ability to remix its AI programming Artifacts, a slew of new AI video tools…

We had GPT-4 and Claude 3.5 write it’s own prompts for Runway’s new GEN-3 text-to-video service and it came up with some VERY weird results. Overall, the experiment wasn’t a smashing success in terms of the videos it turned out but it did teach us that the human in the…