YCombinator Goes AI, Drawing to Website, Local LLMs on Your Mac or PC
A bunch of tools that you can experiment with.
Morning y’all!
Hopefully your eyes aren’t hurting because you played it smart by looking at the eclipse with glasses but apparently there are many who didn’t get the memo. It does make me wonder what folks did hundreds of years ago where everyone was looking at this stuff for long periods of time without protection.
Well, not our problem I suppose! Hope you had a great Monday and are having an even better Tuesday! An assortment of reads today.
¯\_(ツ)_/¯
— Summer
YCombinator is often a bellwether for what’s the come as tons of money is poured into emerging startups. As you might imagine, there are many in the latest batch — 50% of them — that are working on AI and many of them focused specifically on consumer tech. Stay tuned.
Nathan and Alex have a summary of what’s gone done in April (thus far) and it’s got a few links that I hadn’t read or clicked. They focus on geopolitics, hardware, and research among others and I don’t usually follow the first one much. Might be worth subscribing too as well if you want more content.
BoolVideo allows you to transform a URL into a video with AI voices. Kind of neat to see it work. Give it a try if this is something you’re experimenting with.
Concepto allows you to design and code websites with just a simple drawing. This is useful for me since I do most of my work in my notebook and transforming those low fidelity ideas into real sites is definitely interesting. I’m not sure if this would save me much time (on the backend) since I’d have to reformat most of what I’m doing but it might be a good starting place.
Wingman is an open source project to run LLMs locally on your Mac or PC. No code or interfacing with terminals! Wow!
Powernote is voice to text, essentially, but with summaries powered by AI. A simple concept but I know some folks in my life who swear by voice notes and who dictate them all the time. It’s not for me but maybe for you?
Someone shared this on Twitter and I was definitely impressed by the quality. Built by Alibaba’s multimodeal model called EMO. One input image and one audio source. That’s all. Things are getting scary accurate.
Morphic is an AI-powered answer engine with a generative UI. It’s open source to boot. It’s got a decent tech stack too.
Deepseek is another open source project that’s a research agent designed as an internet scale retrieval engine. Quite impressive!
Have you heard of Rooms? It’s an interior decorating app where users can build and code 3D rooms and mini-games. I’ve heard that people are really enjoying it so I’m going to try it this week.
Fixblur is precisely what you think it is. Give it a try. I do love simple apps that solve obvious problems for users.
And that’s it my friends. Have a great one!
※\(^o^)/※
— Summer