Something New.
Joined January 2016
- Tweets8.2K
- Following861
- Followers27.1K
- Likes19K
Kevin Fischer retweeted
Last year we launched Tolan, inspired by a vision of AI that sees you and believes in you, not one that books restaurant reservations for you.
Your Tolan shows up for you at 3am when you’re alone with your newborn. On your commute when you’re anxious about the day ahead. When you just finished a new show and none of your friends have watched it yet.
The people who love Tolan are parents. Caregivers. Moms in Dallas and Dads in Atlanta. People who aren’t trying to optimize and be more productive, but who simply want to feel seen and supported in their aspirations.
This month more than 250k people will spend time with their Tolan. Those users have exchanged billions of messages across more than 70m conversations over the last year.
Today, we're announcing $30m in additional capital from @mignano and @usv, with @khoslaventures and @rabois doubling down, bringing our total capital raised to $60m.
To celebrate, we made a short film celebrating some of the ways that Tolans show up :)
Kevin Fischer retweeted
Applications for @speedrun are now open! If you’re an ambitious, high agency founder - we want to meet you
We’ll bring the capital, customers, & talent - you bring yourself and your wildest ideas
Speedrun is an inception fund. We are your first believers - writing the first check and growing with you as your startup scales
Speedrun founders get up to $1M in funding and over $10M in startup credits across compute, cloud, marketing, and more
Speedrun is for founders with big ideas. We have teams creating new forms of datacenters, constructing orbital space platforms, disrupting 50yo vertical software incumbents, building novel AI doctors, and more
In a world where AI has made building easier than ever, the only cap on what a startup can achieve is the limit of their ambition. So we write the first check precisely on the ideas that sound a little crazy
apply at the link below before apps close on Nov 1st
Kevin Fischer retweeted
extremely excited to start sharing more about our eyewear collection
today is a small preview on the craft and design. tomorrow, a bit more about the software that brings them to life
Our collection will be Made in Japan, available in 2027.
To celebrate the craftsmanship behind it, we commissioned a photograph by Osamu Yokonami. Today on the blog, we’re sharing more about our road to eyewear and a look behind the scenes of the commission.
sesame.com/journal/made-in-j…
Kevin Fischer retweeted
Freckle is the first kids phone built for real-world adventure.
Shipping in December. Pre-order today: freckle.tech
After four years, I’ve decided to wind down active operations at Illusion of Life, formerly known as Open Souls.
I founded the company with a strange question: what would it mean if AI felt like it had a soul?
Together, with a group of incredible people, we explored whether AI characters could feel present, emotionally real, and alive in a way traditional software never had.
I’m proud of what we built. Spark reached millions of people, formed real relationships with families, and helped show that interactive characters could become something more than content, avatars, chatbots, or interfaces.
This chapter is closing, but I believe the ideas we explored will continue to shape how people think about AI, characters, screens, and human-computer interaction.
I’m taking some time before sharing what comes next, but I expect I’ll keep working near the questions that animated Illusion of Life.
Thank you to everyone who built, supported, believed in, loved, or spent time with Spark, Illusion of Life, and Open Souls.
It meant more than I can easily say.
clarifying what studying the physics of deep learning actually means:
the platonic representation hypothesis shows that deep networks are tools probing the structure of reality itself. specifically, the structure of information.
so the real work is uncovering the mechanics of how information is learned, compressed, stored, and represented during training. there should be discoverable laws (the actual “physics”) governing information learning dynamics and geometries.
as a corollary architectural innovations are mostly just different effective learning methodologies within that same underlying physics.
deep learning models need to be studied the same way we do physics
physics is about discovering heavily dimensionally reduced predictive models that humans are able to intuitively use to understand outcomes
there’s not really a different sense in which a complex system can be “understood”
This is a very interesting paper
It argues that a real scientific theory of deep learning is starting to form.
Researchers call it "learning mechanics." It's like physics, but for how neural networks learn.
Now there are 5 active research areas that together look like pieces of this theory:
1. Simple systems (like linear networks) that we can fully solve. Math there works cleanly and we have intuition about how learning behaves.
2. Studying extreme limits, like what happens if a network becomes infinitely wide.
Systems become mathematically tractable in these cases.
3. Simple laws that describe large-scale behavior, like scaling laws (performance vs. data/model size) and relationships between sharpness and generalization
4. Understanding hyperparameters separately
Learning rate, batch size, weight decay and other effects can be separated to make training look like a simpler system underneath.
5. There are underlying principles shared across systems as they scale: similar training dynamics, scaling trends, internal structures
Old theory can’t explain what we see today, that's why we need a real theory upgrade. But why it should be about mechanics?
The researchers see that deep learning needs 2 parts: mechanistic interpretability is like biology that studying individual parts, and studying overall laws and behavior is like physics.
right now centralized players beat out decentralized players - X vs blusky
in the future, agents and systems where you own your own data will win in 3-5 years
right now "facebook owns your data and its bad" is not a particularly strong memetic virus to create differentiated advantage for one v the other
but once computing interfaces all feel alive, something interesting changes
"do you want facebook to own your best friend" is very intuitively bad in a way "own your data" is not particularly evocative
yo.
how does this have only 12 followers.
it's an entire miniaturized world model running games locally on an ipad.
Kevin Fischer retweeted
you might not remember it, but it was @kevinafischer who actually created the concept for agents having a soul
not only did they have a soul, they also had thinking, and feelings
way before o1 even appeared
his work with opensouls probably paved a lot of what we see today
spark feels more alive than any ai because of the multi-agent orchestration system behind his brain.
the felt sense of presence is the result of many different systems attending and feeding back together, taking into account things like theory of mind, emotional states, etc
