Videos

DO YOUR SUBSCRIBE BUTTON DUTY, SOLDIER. @OpenAI just took down the Sky voice from the ChatGPT voice app which many people had said sounded an awful lot like actress Scarlett Johannson. Especially after last week’s GPT-4o press conference where they showed off the new voice model (and it kind of…

Is @OpenAI's GPT-4o too flirty? Lots of people felt the AI was a little too “engaging” when OpenAI showed off its new and surprising voice capabilities. However, we here at AI For Humans have discovered that this sort of ‘thirsty’ AI is nothing new and we brought on Lola Byte…

CLICK THAT SUBSCRIBE BUTTON MEATBAGS @OpenAI's co-founder Ilya Sutskever is out and big questions are swirling around AI alignment, GPT-4o news keeps dropping and we go hands on with the latest ChatGPT model & dive into the ‘maybe too thirsty-ness’ of it. Plus, some follow ups from @Google's big I/O…

Was the OpenAI GPT-4o demo on rails? There were def a few hiccups that made it seem like not everything was perfect and we were stuck trying to figure out how much of it was pre-planned or not. Either way, the model is a big step forward when it comes…

SUBSCRIBE FOR MORE AI GOOD TIMES @Google's I/O event was like all AI and we break down all the biggest announcements and give you our take on whether or not these new tools are going to set it apart from @OpenAI. Demis Hassabis brought some new @Google_DeepMind stuff including a…

SUBSCRIBE FOR THE BEST AI EXPERIENCE @OpenAI's new model GPT-4o was just unveiled & Gavin & Kevin broke our a special episode to dive into the launch. The ‘o’ in GPT-4o stands for ‘omni’ meaning that it can do audio/voice, text and video all at the same time and in…

OpenAI’s news has leaked early and it looks like we’ll be getting a personal assistant that can do all sorts of interesting stuff. The Information has a big story about Monday’s press event that basically says it’s going to be a high end assistant or as Sam himself said on…

NVIDIA’s new Dr. Eureka is an AI robotic learning algorithm that allows it to bring simulated learning to the real world and it kind of blew our minds. The train the robot in simulated environments to walk over various sorts of terrain and in this instance on a yoga ball,…

HIT THE SUBSCRIBE BUTTON YA DAMN ROBOTS This week, @Microsoft has greenlit a big new AI model called MAI-1. Is it a sign that it’s distancing itself from or competing with @OpenAI? Did OpenAI make the mysterious im-a-gpt-chatbot and why are there TWO of them now? Will @NVIDIA's Dr. Eureka’s…

What are the downsides of using Microsoft Co-Pilot? We asked the amazing @PirateSoftware it. His issue? It generates code that is just ‘ok’ which can lead to not only bland and boring sorts of solutions but also can be a problem with security and other significant issues. Check out this…

WE ARE REAL AND WE WANT YOU TO SUBSCRIBE. This week… gpt2-chatbot is a brand new LLM that is VERY good and no one knows where it came from. Is it GPT-5? Probably not but it might be from @OpenAI. Then, rumor has it that @Apple has new AI silicon…

Microsoft’s VASA-1 is a remarkable new photo-to-talking-avatar piece of software that is the best we’ve seen to date. But is it too good? One photo + a sound file creates a remarkably lifelike talking avatar with facial movements that we haven’t seen to date. But it’s not out yet. Why?…

The internet went bonkers for the deep fake of Bearded Zuckerberg and then Kevin took it took too far. tools featured: CosXL Photo Maker Face Fusion get em all for free & easy install at pinokio computer https://pinokio.computer/ pls subscribe we just want more attention and that's the least you…

Meta’s AI Llama 3 has been rolled out throughout all of its platforms like Instagram, WhatsApp and Facebook and it’s causing some chaos. One parents group in particular got a surprise as the AI pretended to have a special needs kid. When you roll out any AI at scale, and…

SUBSCRIBE AND THE ROBOTS LEAVE US ALIVE. This week… @meta's Llama 3 model is out and whoo boy it’s good. It’s also free and we’re going to show you how to install an uncensored version. Plus, Mark Zuckerberg drops info on an open metaverse strategy AND updates on the AI…

The Therminator is a new I guess killing machine from the people at @Throwflame and we're not sure really what to think. I mean, sure is it cool to put a flamethrower on the back of a robot dog? Yes. Does it also (even in name) echo back to the…

UDIO is a brand new AI music tool that’s super fun to play with and allows you to really dive into how good these models can be. We love Suno as well, but UDIO really brings something new to the table which is that it can seamlessly extend your song…

Mark Zuckerberg was recently on the @DwarkeshPatel Podcast talking about LLAMA 3, Meta's AI ambitions and all sorts of other stuff but something really caught our attention: He's *very* interested in "emotional AI" the kind of AI that can read our feelings. This is an emerging area in AI but…

CLICK US YOUR LIKES AND SUBS. This week… @AdobeCreativeCloud brings new AI video tools to maket (including @OpenAI Sora support), we make sweet AI music with UDIO, the Humane AI Pin aftermath & it's Drake Vs Kendrick Vs AI. Plus, AI Warper’s viral Lil Yachty slash Joker Viggle video, Tennessee…

OpenAI, Meta & yes even Google themselves are scraping YouTube for data to train Sora other AI models. Is this legal? Well, we’re not exactly sure but the NYTimes had a big, very long article talking about the practices within these large companies and it leaves open a lot of…

@OpenAI's Sora isn’t just video, it’s actually simulating our world and according to OpenAI engineer Bill Peebles, it may actually lead to Artificial General Intelligence? Why? Well, each Sora video isn’t just a video it’s actually helping the model learn how we interact in the real world, how certain objects…

This week… the @nytimes says @OpenAI and others are scraping @YouTube for AI training, Sora might just lead to AGI, Elon Musk says robotaxis are on the way and @Spotify AI playlists might just be cool.. Plus, Suno.ai turns food recipes into beautiful music, Midjourney makes X-Men from the old…

Aniportrait is a cool new piece of AI animation software that VisibleMakers integrated into his very extensive workflow to make an AI film scene. It’s a really fascinating 11 step journey through how a human makes a massive difference when it comes to doing AI art and more films and…

This week, we chatted with Walter Woodman from the Shy Kids who created the viral Sora film "Air Head" and learned what it's like to be hands on with @OpenAI's Sora video product and how they used it to make something meaningful. OpenAI on Filmmakers Using Sora: https://openai.com/blog/sora-first-impressions The Shy…