That intro makes me want to cry, because she's describing what the hope was for augmented reality before Zuck and Tim and every web developer looking for the next step on their career ladder stuck their grubby hands into its chest and squeezed its aorta shut. AR wasn't floating screens. It wasn't hackneyed VR. It was digital bits and bobs integrated into your physical space, onto your physical objects. You know, augmenting them. Like this, but with glasses or a headset instead of a projector. And (particularly at the beginning) not so much hyper-optimized for enterprise productivity, as for doing something small and interesting and maybe a bit useful.
We came so damn close, and it's been ~10 years. Maybe this gets people's imaginations going again. Get us ready to take that, er, magic leap forward.
i dropped a mail but would love to add regarding AR that while its correct to integrate real world objects as inputs, using the same as output messes with your brain and skips the part where you transition a user from his known workflow to a more integrated one.
So instead of overlaying and augmenting the room the user already has mapped to functions i propose first extending the room which can be done via 2 projectors and passive 3D more easily and will offload all spatial senses to a new area (z all far aquarium like) that a user can see as something added and new and not something breaking his current status quo.
A large second virtual room on the wall to work with is around 1000€ for 2 projectors, silver screen and some lenses. Growing new neurons has an "aha" effect on the user while forcing him to rewrite them will leave them with pain and force a defensive reflex - you don't want that while proposing new concepts!
If you like overlaying stuff a passive 3DTV somewhere in the room or on your desk can do a floating object easily for under 300€ (z all near) while not messing with more than this small experimental area of the room. Black backgrounds work great.
This allows the user to smoothly transition from his used-to workflow to a more integrated one all while just wearing 10g set of glasses - while we wait for affordable AR glasses.
Skipping this part breaks the users known working toolset - you can do so maybe (if ever) AFTER he gets comfortable with the new toolset. Better analogy: Imagine being a teacher and i propose a camera that will turn what you write on the whiteboard into an animated explanation (graph visualization of your formula). Now your whiteboard still works! Nothing is broken. Maybe for stuff you already now the extension works just fine and your students see the graph immediately on ANOTHER screen area. That's great and added value. Because even if you didn't learn some syntax your whiteboard will always work as before and you can just draw the graph yourself or turn the second screen off then go on with your life and continue the lesson.
Now maybe if you got comfortable with that you will allow the second area screen on the whiteboard itself next to your drawn formulas (or typehinting them even;) ). But not during learning how it works.
I took the time to read through your software stack. While i understand the concept makes sense (i propose similar self hosted solutions and applaud the math proven security aspect) i want to show you why you lack people leaning more into it:
As the 7 parts already hint there's a limit to what people can comprehend in any amount of time so while its what you do and know and fills your millers number this is not the same for others.
Trying to get customers, any customers, will teach you the customer has obligations to fulfill and he's looking for a way to go from his A to his imagined B. Now you maybe can offer some B' that has parts of your solution, but that can only be 1 or 2 parts of your stack at maximum and bend his imagined B at a very tiny amount or he will not accept it.
Because the rest of his millers number (5 or whatever intellect he shows, maybe more) is already filled with family, his boss, bottom line finances and time. Also there's always pain when you change some thinking - even if your solution is better. Most people will look for a quick fix then move on to their next immediate problem.
So maybe with some luck after handing over 3 or 4 B' to them that all interconnect you can propose a greenfield solution like yours that's now just 2 steps away.
But going there is transition and not greenfield - the customer has a somewhat working system already and completely replacing this will not make him happy.
Maybe he has a working Active Directory? Or just an apple time capsule whatever? Gmail business? Can you transition from that for the same price and pain your competitors offer as an "immediate" solution?
I do not get the "against usefulness" portion. The article still discusses the projects she deems "useless" in relation to their future potential usefulness.
Deep, paradigm-shifting research often needs decades of deep thought and experimentation, which will not yield weekly, monthly or even annual demonstrations of usefulness to customers or shareholders.
She writes as a VC (I think?) where her job is to allocate capital for shorter-term "useful" (read: profitable) outcomes.
My point is that something is not useless if it has the potential for future use. I would, for example, not call the example in the article (in any stage of its development) useless because it has the potential. Same for any possible paradigm-changing work. If it is proven (in the strictes sense) to not work, then it could be deemed useless.
Now if something is profitable is a whole different thing.
The thing is you don't know that it's got actual future usefulness. You might think or hypothesize that it does, but for various reasons it may turn out to be a dead-end.
Almost every endeavour has non-zero potential future utility - even as a counterexample or lesson to learn from. I think that definition of "useful" is probably too broad to make an argument against, and so not a useful definition.
And you have spotted the early stage opposrtunity most VCs fail to fund: paradigm shifting work often looks like infrastructure, very broad and general-purpose. It can eventually lead to trillions of dollars in wealth creation (the Web, Linux, smart contracts, etc) but the first movers may capture 5% of the ecosystem’s value. For a VC, that’s a bigger bet that can change the world a lot more than, say, the next short term memecoin sold to the greater fool for 50x or hype app dumped on the market for 100x.
Paradigm work has always run on patrons. A new paradigm is useless at the beginning, by definition, so markets have nothing to price. The first personal computers were toys. So was the early internet. The people who fund this stage are buying a position in the future before anyone can price it.
Well, some of us are drawn to this world-changing work (Vint Cerf who’s retiring this week, Linus Torvalds, Tim Berners-Lee, Vitalik Buterin, Jimmy Wales, Igor Sysoev etc etc) and it runs most of the Internet. But, will it get funded?
There’s a tactile element to understanding your assets, supply lines, air/sea/land bridges, time to rendezvous, enemy dispositions, etc that could benefit from a richer visual canvas.
How to take that physical state, encode it so it can be replayed/evolved/disseminated into taskings.
At the moment it’s all spreadsheets, white boards and staff officers.
While I haven't done a lot with the idea (one month of concentrated work a few years ago, before LLMs), my idea is: the stylus needs to become way more digitized. Keyboards and trackpads are fine, but a computer that understands intent based on how we squiggle, that will help us to get out of our confined way of thinking.
I used to have hacker news web pages that I scribbled upon with a stylus.
Web pages should feel more like paper.
But I'm currently not in a position to work on something like this because I have a pay check to earn. The pay checks that come out of university aren't good enough.
A fascinating read. Currently, what are the best AR Glasses that could replicate this setup, but just for me ? I am not a developer but just out of interest, I feel like there are interesting things at play here.
The product is open source, I liked it in the article, you can set it up yourself or get help from Omar, no glasses needed. They also have a small portable hardware device, the one I was holding in my hand, happy to help lmk oana@motiveforce.ai
A recent discussion here[1] also spiked my interest in "physical computing"
I'd love to get on board with people working / brainstorming / exploring new possibilities. I'm very inclined for education, I think the potential is tremendous, but could have plenty other use cases
That's DynamicLab, which has been around for years. Like most 2D programming environments, it runs into a scaling problem. At some point, the diagram becomes too big. This is why complex logic is designed in VHDL rather than with really big schematics.
There was an era of giant books of schematics. It sucked.
that's why i extended diagrams to n dimensions, the structure can grow while each expertise keeps its own 2D POV not touching the flat diagram of others: https://krei.se/Editor/nodeEditor
With that you can just draw your idea in a 2D plane and if you hit the space available wall change to another 2D plane drag stuff around differently but this will not mess with the original 2D plane. When you want to discuss with other POVs rotate to a shared angle.
relatable. my AR apps (visionOS) are also in prototype/exploration/useless stage despite successful iOS/Android/web, pretty much like whole platform except few usecases like movies (wall-gardeneed). doing literally nothing and looking around in visionOS is one of my favorite features (tastement to good technology, lack of market, and walled garden, which is a shame).
Yes sadly it’s no longer a thing, Bret is not a fan of LLMs. I’m trying to find a “director” to his “creative” someone who can help raise money and energize him. I don’t even know Bret personally but I want to see what comes next. Share if you know people oana@motiveforce.ai
correction, maybe it's still alive? a friend shared https://x.com/lukexi don't know for a fact but i very much admire Bret and the entire team and I want them alive and working and I am happy to contribute any way I can
You hand a webcam and a projector over it, run image detection, link by physical adjacency.
I was adjacent to a BigTech project working on a variant of this about 15 years ago - had the projector and camera both embedded in a smart lightbulb. Unfortunately never saw the light of day
That intro makes me want to cry, because she's describing what the hope was for augmented reality before Zuck and Tim and every web developer looking for the next step on their career ladder stuck their grubby hands into its chest and squeezed its aorta shut. AR wasn't floating screens. It wasn't hackneyed VR. It was digital bits and bobs integrated into your physical space, onto your physical objects. You know, augmenting them. Like this, but with glasses or a headset instead of a projector. And (particularly at the beginning) not so much hyper-optimized for enterprise productivity, as for doing something small and interesting and maybe a bit useful.
We came so damn close, and it's been ~10 years. Maybe this gets people's imaginations going again. Get us ready to take that, er, magic leap forward.
I just want to know how many HP I have left
Glad to hear I was able to spark some emotion. I’d love to chat oana@motiveforce.ai
i dropped a mail but would love to add regarding AR that while its correct to integrate real world objects as inputs, using the same as output messes with your brain and skips the part where you transition a user from his known workflow to a more integrated one.
So instead of overlaying and augmenting the room the user already has mapped to functions i propose first extending the room which can be done via 2 projectors and passive 3D more easily and will offload all spatial senses to a new area (z all far aquarium like) that a user can see as something added and new and not something breaking his current status quo.
A large second virtual room on the wall to work with is around 1000€ for 2 projectors, silver screen and some lenses. Growing new neurons has an "aha" effect on the user while forcing him to rewrite them will leave them with pain and force a defensive reflex - you don't want that while proposing new concepts!
If you like overlaying stuff a passive 3DTV somewhere in the room or on your desk can do a floating object easily for under 300€ (z all near) while not messing with more than this small experimental area of the room. Black backgrounds work great.
This allows the user to smoothly transition from his used-to workflow to a more integrated one all while just wearing 10g set of glasses - while we wait for affordable AR glasses.
Skipping this part breaks the users known working toolset - you can do so maybe (if ever) AFTER he gets comfortable with the new toolset. Better analogy: Imagine being a teacher and i propose a camera that will turn what you write on the whiteboard into an animated explanation (graph visualization of your formula). Now your whiteboard still works! Nothing is broken. Maybe for stuff you already now the extension works just fine and your students see the graph immediately on ANOTHER screen area. That's great and added value. Because even if you didn't learn some syntax your whiteboard will always work as before and you can just draw the graph yourself or turn the second screen off then go on with your life and continue the lesson.
Now maybe if you got comfortable with that you will allow the second area screen on the whiteboard itself next to your drawn formulas (or typehinting them even;) ). But not during learning how it works.
Certainly. I will shoot you an email with this username in the heading.
> ...two researchers who worked at Dynamicland, Bret Victor’s lab in Oakland.
> Everyone is building the same thing, funded by the same people, using the same words. Where are the independent thinkers?
> [JP] said a name: Folk Computer. In New York.
Who are ... building the same thing as Bret. So where are these independent thinkers?
I’m right here
Been building about 20 interconnected pieces of a new software stack designed to help humanity.
Most VCs wouldn’t fund it.
Most investors wouldn’t understand it.
They prefer small apps. YC’s startups used to be called “feature, not a startup”.
Now with AI it has become easier — the work I and my small team that I’ve been paying for 10 years has been greatly accelerated.
But still. Most people on HN will just deride it. Few would look deeply. Here, I will share a link:
https://safebots.ai/ETHOS.html
I took the time to read through your software stack. While i understand the concept makes sense (i propose similar self hosted solutions and applaud the math proven security aspect) i want to show you why you lack people leaning more into it:
As the 7 parts already hint there's a limit to what people can comprehend in any amount of time so while its what you do and know and fills your millers number this is not the same for others.
Trying to get customers, any customers, will teach you the customer has obligations to fulfill and he's looking for a way to go from his A to his imagined B. Now you maybe can offer some B' that has parts of your solution, but that can only be 1 or 2 parts of your stack at maximum and bend his imagined B at a very tiny amount or he will not accept it.
Because the rest of his millers number (5 or whatever intellect he shows, maybe more) is already filled with family, his boss, bottom line finances and time. Also there's always pain when you change some thinking - even if your solution is better. Most people will look for a quick fix then move on to their next immediate problem.
So maybe with some luck after handing over 3 or 4 B' to them that all interconnect you can propose a greenfield solution like yours that's now just 2 steps away.
But going there is transition and not greenfield - the customer has a somewhat working system already and completely replacing this will not make him happy.
Maybe he has a working Active Directory? Or just an apple time capsule whatever? Gmail business? Can you transition from that for the same price and pain your competitors offer as an "immediate" solution?
The summary?
> “I don’t need the monitor anymore.”
Yeah, you just need a complicated setup of machinery that observes the room and multiple conventions on content tagging, etc.
I do not get the "against usefulness" portion. The article still discusses the projects she deems "useless" in relation to their future potential usefulness.
There appears to be a silent "immediate" in this version of "useful".
Against "only immediate/certain usefulness".
Deep, paradigm-shifting research often needs decades of deep thought and experimentation, which will not yield weekly, monthly or even annual demonstrations of usefulness to customers or shareholders.
She writes as a VC (I think?) where her job is to allocate capital for shorter-term "useful" (read: profitable) outcomes.
correct, here about my firm - 2nd quarter of being alive and it's just me https://motiveforce.ai/
My point is that something is not useless if it has the potential for future use. I would, for example, not call the example in the article (in any stage of its development) useless because it has the potential. Same for any possible paradigm-changing work. If it is proven (in the strictes sense) to not work, then it could be deemed useless.
Now if something is profitable is a whole different thing.
The thing is you don't know that it's got actual future usefulness. You might think or hypothesize that it does, but for various reasons it may turn out to be a dead-end.
Almost every endeavour has non-zero potential future utility - even as a counterexample or lesson to learn from. I think that definition of "useful" is probably too broad to make an argument against, and so not a useful definition.
Congratulations on putting it together!
And you have spotted the early stage opposrtunity most VCs fail to fund: paradigm shifting work often looks like infrastructure, very broad and general-purpose. It can eventually lead to trillions of dollars in wealth creation (the Web, Linux, smart contracts, etc) but the first movers may capture 5% of the ecosystem’s value. For a VC, that’s a bigger bet that can change the world a lot more than, say, the next short term memecoin sold to the greater fool for 50x or hype app dumped on the market for 100x.
Paradigm work has always run on patrons. A new paradigm is useless at the beginning, by definition, so markets have nothing to price. The first personal computers were toys. So was the early internet. The people who fund this stage are buying a position in the future before anyone can price it.
Well, some of us are drawn to this world-changing work (Vint Cerf who’s retiring this week, Linus Torvalds, Tim Berners-Lee, Vitalik Buterin, Jimmy Wales, Igor Sysoev etc etc) and it runs most of the Internet. But, will it get funded?
Here’s mine:
https://safebots.ai/about
See something here re mission ops planning.
There’s a tactile element to understanding your assets, supply lines, air/sea/land bridges, time to rendezvous, enemy dispositions, etc that could benefit from a richer visual canvas.
How to take that physical state, encode it so it can be replayed/evolved/disseminated into taskings.
At the moment it’s all spreadsheets, white boards and staff officers.
Thinking cap on - thanks for the post!
While I haven't done a lot with the idea (one month of concentrated work a few years ago, before LLMs), my idea is: the stylus needs to become way more digitized. Keyboards and trackpads are fine, but a computer that understands intent based on how we squiggle, that will help us to get out of our confined way of thinking.
I used to have hacker news web pages that I scribbled upon with a stylus.
Web pages should feel more like paper.
But I'm currently not in a position to work on something like this because I have a pay check to earn. The pay checks that come out of university aren't good enough.
A fascinating read. Currently, what are the best AR Glasses that could replicate this setup, but just for me ? I am not a developer but just out of interest, I feel like there are interesting things at play here.
The product is open source, I liked it in the article, you can set it up yourself or get help from Omar, no glasses needed. They also have a small portable hardware device, the one I was holding in my hand, happy to help lmk oana@motiveforce.ai
A recent discussion here[1] also spiked my interest in "physical computing" I'd love to get on board with people working / brainstorming / exploring new possibilities. I'm very inclined for education, I think the potential is tremendous, but could have plenty other use cases
[1] https://news.ycombinator.com/item?id=47747770
Very good article from Oana - she is the real help for founders
That's DynamicLab, which has been around for years. Like most 2D programming environments, it runs into a scaling problem. At some point, the diagram becomes too big. This is why complex logic is designed in VHDL rather than with really big schematics.
There was an era of giant books of schematics. It sucked.
that's why i extended diagrams to n dimensions, the structure can grow while each expertise keeps its own 2D POV not touching the flat diagram of others: https://krei.se/Editor/nodeEditor
With that you can just draw your idea in a 2D plane and if you hit the space available wall change to another 2D plane drag stuff around differently but this will not mess with the original 2D plane. When you want to discuss with other POVs rotate to a shared angle.
relatable. my AR apps (visionOS) are also in prototype/exploration/useless stage despite successful iOS/Android/web, pretty much like whole platform except few usecases like movies (wall-gardeneed). doing literally nothing and looking around in visionOS is one of my favorite features (tastement to good technology, lack of market, and walled garden, which is a shame).
I agree. I occasionally use the fishing app
This reads like Dynamicland is no longer a thing. Their website seems to end in 2024. Anyone have any idea what Bret Victor is up to these days?
Yes sadly it’s no longer a thing, Bret is not a fan of LLMs. I’m trying to find a “director” to his “creative” someone who can help raise money and energize him. I don’t even know Bret personally but I want to see what comes next. Share if you know people oana@motiveforce.ai
correction, maybe it's still alive? a friend shared https://x.com/lukexi don't know for a fact but i very much admire Bret and the entire team and I want them alive and working and I am happy to contribute any way I can
Showed some love on his https://seastone.bandcamp.com maybe he'll reply
How do you plug a keyboard into a sheet of paper?
You hand a webcam and a projector over it, run image detection, link by physical adjacency.
I was adjacent to a BigTech project working on a variant of this about 15 years ago - had the projector and camera both embedded in a smart lightbulb. Unfortunately never saw the light of day
probably wireless keyboard and routing it based on physical proximity
This hits home - I wrote you an email!
I feel like VCs are less relevant than ever.
It’s nice to see gatekeepers vanish to irrelevance.
The environment allows for separation of the useful good ones from the "only in it for the cash" ex-investment banking types ^_^
[dead]