Partly Sunny with a Chance of AI

HuggingFace updates their open LLM leaderboard, and the model that's better than Facebook/Meta's Llama 3-70B model is from Alibaba, called Qwen2-72B-instruct. Check out our post to see how they've updated the evaluation criteria: https://lnkd.in/eT47zjQZ

Runway's Gen-3 model to text-to-video generation is available to subscribers. We tested it with some prompts and it's impressive, but there's a learning curve to steer it with a lot of trial and error. Check out our free post to see the results: https://lnkd.in/eEh-Wy-H

Google's smallest AI model is coming to browsers. Currently limited to developers. Check out the details in our post: https://lnkd.in/efGen8TZ

Low-code and no-code are a temporary labels we apply to what's recently accessible that needed someone to write code to build. As software is abstracted, we drop the adjectives and call the things for their function. To understand the maturity and prevalence of a given tool, see which ones we call no-code vs. which ones we simply call by their function. When was the last time you heard someone describe Gmail as no-code email? It's simply email. There's also a distinction between what needs code vs. what is technical, don't confuse the two! Read about how we define these terms and distinguish them in our latest post: https://lnkd.in/eqsPi23W

Figma announces new AI features supporting their core workflows. Check out our blog post to learn about the big and small features and the implications for designers: https://lnkd.in/eWCYBkTA

If generative AI is a platform shift, what's the fundamental skill of our era? We reflected on this topic in our post to identify several key skills in relation to where generative AI excels now. https://lnkd.in/enncGT3H

EvolutionaryScale announces their latest model, ESM3, that can generate protein structures from text prompts. Check out our blog post (requires premium subscription) on how ESM3 works and the implications on the life sciences: https://lnkd.in/ebgbXAd2

From text to user interface in seconds. Galileo's model creates mobile and web interfaces based on a text or image prompt. Check out our free article to learn more about it: https://lnkd.in/gPHwB_GJ

Google Deepmind shares progress on their video to audio model. It's pretty good, and another piece to generating realistic videos. Check out the blog post and samples in today's newsletter: https://lnkd.in/emBhWyRf

How does generative AI change how we create and value software? A few takes on the end of software to disposable software. Claude 3.5 Sonnet does make it seem like it's more of a reality. Check out the post here: https://lnkd.in/eqDHYz9J