
consumer ⚭ hw @opalelectronics
sf (prev ldn, sg)
Joined February 2012
- Tweets280
- Following908
- Followers831
- Likes11.6K
solderneer retweeted
When I saw the first fMRI-to-image models in 2022, my first question was “is this real?” My second was “how far can we take this?”
So gratifying to finally share some of what we've made over the past three years. Very excited for the role BCI will play in helping exponentially more powerful AI align with humanity’s best interests.
Introducing AIR-O, a model that decodes what a person is attending to directly from brain scans.
Using only noninvasive brain data, we can generate an image from scratch in real time that depicts where a person is focusing their attention.
This is the first of a new generation of algorithms that can read the shape of thoughts directly from brain activity.
fable 5.2 is actually an on-par human PCB router but its default out of the box intuition is bad without some prompting
1
1
146
solderneer retweeted
it annoys me so much that media portrays plastic as a purely evil material that we're terrible for still using. replacing your straw isn't going to do anything if hundreds of millions of tonnes are still being produced annually, it just makes you feel better for 'trying'. like concrete or gas, an alternative needs to scale
it's useful to stop and notice that all the properties that made plastic so popular are the same ones that we dislike. you cannot have one half of what the properties enabled without the other. some notes on what's actually hard about replacing plastic: chandhana.blog/p/plastic
realising that having multiple agents for different things (à la grokbot) is a vastly superior product experience then a single super agent (à la dot)
3
108
solderneer retweeted
I'm now building an "improvements" feature into all of my personal vibecoded applications: a freeform text input that's hooked up to an agent that yolo implements whatever is submitted.
It's an interesting experience -- the activation energy barrier to sanding any rough edge or implementing any new feature becomes extremely low. Someone else who uses one of the applications I created has already implemented 150 improvements over the course of a few days. This pattern feels like a new kind of tool -- liquid software? -- and every application that doesn't have this feature now has a slightly rigid and anachronistic flavor. Why can't I just tell it how I want it to be different?
Now, it's not obvious to me how this pattern can or should scale. Does it become unmaintainable after a while? (Maybe model intelligence improvement outstrips my feature appetite such that it's fine?) How does collaborative software work in this model? Can software designed this way be secure and reliable?
Whatever the larger questions about broader deployment, it definitely works great in a personal context. It reminds me a little of the visions of Lisp/Lisp machines, Smalltalk, and HyperCard -- systems that the user can take control of and modify. Those visions never quite worked, though: the vast majority of software today is not end-user modifiable in any useful way. So, while this "liquid" affordance is a very small touch, the way it changes one's relation to the software feels a bit transgressive and empowering in a Lickliderian and Engelbartian way.
automated my work with codex so I can spend more time on human pursuits like trying to find a house in sf I can afford on zillow
1
1
96
make linux beautiful again
Pond, our linux operating system, is now available for anyone to try!
Our updated website includes the installers and a detailed look at the operating system
butterfly.so/pond
6
275
we should explore more alien creatures in robotics. theres such a vast design space of morphologies unexplored by biology.
like why does every home robot have to be humanoid or dog???
MorpHex: The Shape-Shifting Hexapod That Transforms Into a Rolling Ball
via @ctorobotics
#Robotics #EmergingTech #Technology #Innovation
1
10
449


















