Does the world look different to you?

What a week!


New Podcast

Latest podcast with co-founder of Vapi- an AI voice SDK (software developer kit).
This time, it’s great on audio or video
Spotify
Youtube

How didn’t I know that?

Context length or window  is not what you get as an output of an LLM.
Example - Gemini pro having 1 mil context window. That doesn’t mean it’ll output 1 million tokens. It means the volume it can take in. 🤯
This may seem obvious to most of you but it was a turning moment for me.

Rabbit holes I’m going down

How do we preserve the internet? (on link rot)
being an “energy vampire”

what does the future of ai agents look like?

From the community

Summing up what we’re doing…

For Thoughts

My internet friend   sparked a discussion about gatekeeping in the tech industry and the harmful effects of "no true Scotsman" fallacies.

The idea of moving goal posts for definitions hindering
1) progress and 2) participation.
This, for me is right on the money and especially relevant to how “non-technical” people show up to support and scale anything in tech.
Makes me think of the above meme (Andrew Ng- Deeplearning) and leads me to think about  how we can be less gatekeepy.
- meeting people where they’re at
- giving little context breadcrumbs/clues
- learn in public style content (ala Swyx )

OS Section

Grey software is a project I’ve recently found, dedicated to building the open education ecosystem of the future. They have an Open Education Strategy, Open Collaboration Strategy. They’ve got comprehensive modules (like here) to get you started, for example “how to use Markdown”.

Resource Rec

We’re live, baby. The Unsupervised Learning Community Slack Channel kicked off this week- complete with newcomer quick-start guide, project ideas, contributor page + more. Check out our dedicated resource page that has a collection of youtube videos, articles, tools and people to follow with a comprehensive tagging/filter system.

Stay curious,

Renee