Introducing Kino: make videos end-to-end with Claude Code and Codex.
Drop your footage, get highlights, edit together, send links for feedback, and let your agents cook.
AI is becoming a great creative partner. Kino is a new home for your team and agents to collaborate.
Introducing Kino: make videos end-to-end with Claude Code and Codex.
Drop your footage, get highlights, edit together, send links for feedback, and let your agents cook.
AI is becoming a great creative partner. Kino is a new home for your team and agents to collaborate.
Been thinking about this the past few days. Some thoughts
1. How does this fit into our definition of context switching?
2. Is adhd a low-consciousness disorder? Lol
3. Is the experience of focus (and more specifically flow state) “elevated consciousness” ? Maybe the Buddhists were right all along
4. If we can detect / decode what the unconscious processing part of the brain is doing, can we treat attention disorders with some form of non-chemical treatment that trains us to recognize when we’re moving out of focus?
5. I wonder how much of depression is related to this as well and if we’ll ever be able to more objectively diagnose individual sources of depression or anxiety via subconscious decoding
1/ We discovered something totally unexpected about the neural structure of consciousness that overturns decades of thinking on its biological basis and brings us significantly closer to the recipe for generating conscious awareness.
New preprint: (led by @HesseJanis, with @FrankLanfranchi, @_conor_mcgrory, @Yuelin_Shi, and @EngelTatiana)
it is classification all the way down
and now that models are great at training models themselves (especially smaller ones that can be trained off of larger backbones) the bottleneck is creating and cleaning synthetic data
The most obvious cases for a company to fine-tune a model today is anything that looks like "old ml": classification, search, routing, etc.
In the future, taste will be a big area too. Not just "did the model do the thing?", but "did them model do the thing the way I wanted it to?"
Introducing Enzyme: a language for teaching agents what to notice in your Markdown wiki.
You declare how different parts of a knowledge base should be read. Enzyme compiles that lens into questions your agent can use to find the right sources.
You keep talking to your agent normally.
Honestly wow I am so excited for stuff like this especially when it can be done completely locally at same quality level. Will be complete holy grail programming
I've seen people describe Jev as an "AI if statement". But what if it actually WAS an if statement?
Introducing Probably: a programming language powered by Jev:
Jev baked into the language. “feels” asks a question. “match” routes between descriptions. “while” keeps going until something stops feeling true.
This is obviously a toy, but it's fun to think about what something like Jev unlocks. Jev makes the decisions, an LLM does the writing, and a little program ties it together.
let me be more specific because Interact is elusive to many
Interact members have founded: Etched, Graphite, Figma, Cognition, Figma, Scale, Socket, Modal
Also some of the best operators. Start asking around and you'll see for yourself... they're everywhere!