Building a Skrode — initial thoughts
As I get older, I feel my mental faculties slipping. Memory of all types is degrading. My working and short term memories seem to be dodgier than they used to be, and I suspect my long term memory of being almost entirely fabricated. The interesting thing is that I’ve lived long enough to notice the change. I don’t think I have any kind of condition or health problem, it’s just being 44.
As this has been progressing, I’ve been wondering about addressing this slippage with tech. For instance, as a youngster I used to code on tiny screens, but nowadays I use as many large monitors as I can get; they help me retain context, taking the load off my working memory. So tech can definitely help if you use it well.
One of the things I’m thinking about primarily is adopting a prosthetic memory.
From observing people who are quite a bit older, it’s clear this all progresses, and that past a certain point it’s so hard to learn new things that a tech based prosthetic memory, for instance, might be impossible to adopt. Once you hit your 70s or 80s, you really want to be working with what you already know.
So if I’m going to adopt a prosthetic memory I need to start now, make it a habit.
Of course we all have prosthetic memories now to some extent. If you’re a smartphone user who auto uploads your photos to Google Photos, you’ll know what I mean. Even just using Facebook is like this. These platforms pop up and tell you what you were up to 4 years ago on this day, things like that. And you can go back and search through your history. Very cool stuff.
But I’d like to go further than that.
In Vernor Vinge’s book A Fire Upon The Deep, the alien Skroderiders are intelligent plants that were uplifted by giving them electro-mechanical “skrodes”, kind of wheelchairs with intelligence augmentation, most significantly prosthetic short term memory and databases of information, which their minds otherwise lacked.
I don’t need the wheelchair (yet!), but I’m feeling a bit more like a Greater Skroderider every day. So it makes sense to think about building myself a Skrode.
What could the elements of a Skrode be?
- Ideally everything you see and hear should be captured and searchable. I saw a reference to a startup with software that captures everything you do on a computer, I think by OCRing the screen, but I can’t link it because I can’t remember any details :-) .
- Also, being able to take lifelogging photos and videos and have them automatically tagged and searchable would be great.
- Metrics: I’d love to be able to capture time-series measurements in a really generic way. Think of the kinds of things that a fitbit captures automatically (pulse readings? step counts?). But also things like how much coffee I drank today, or that I’m having a headache right now.
- Analysis: Some kind of analysis of all the data being gathered, looking for higher level insights??
- Dashboard?: Visual display of metrics, analysis, associated information that relates to whatever is happening in the moment. This might need to be grouped into contextual profiles; eg: I might want a totally different dashboard when I’m coding to when I’m walking around to when I’m writing.
- Intelligent agent interface: Alexa or Google Assistant or similar should be able to access all this stored memory, using strong custom skills / actions for this task, and provide me a high level way of interacting with it.
- Realtime Associations: Some combination of the analysis layer and intelligent agent interface should be pushing notifications to me in real time based on context. eg: if I’m reading an article, I should also be getting a stream of related things that I’ve read before on this topic, and maybe new things that I haven’t seen but might interest me based on the content of my prosthetic memory and the content of what I’m reading.
- Knowledge Model: The skrode should try to model what I know. If it can do this, it can notice when I’m trying to understand something that’s beyond me because of some kind of knowledge gap, and push me in the right direction. It could also use this to help me learn things. If I tell it I’m trying to learn about, say, molecular biology, and it has a model of what related / required things I know, it can provide me with a suggested path for learning, tailored to my specific learning up to now. It could refine the model by probing me with questions. It could implement spaced repetition, bringing back material periodically to reinforce learning.
And a lot more things. The list above is a bit rambling; this is a new idea, so I haven’t thought it through well yet. It’d be great to boil it down to a basis set of functions. My instinct is that the skrode could be a coherent platform that could be customised, but the basic tools of which could be used by anyone.
What might the hardware / UX aspects look like?
- It should work on all the devices I use, and sort of follow me from device to device, capturing everything, providing the services as above.
- It could use a heads-up display, like Google Glass. I’d love that. I think the tech’s not quite there yet though. It might do just as well by just appearing on whatever screen(s) I’m working on and displaying there.
- A conversational interface would also be excellent, like alexa/google now/google assistant/siri/cortana. I could wear a bluetooth headset a lot of the time, or it could just turn up on devices as above. It doesn’t necessarily need to be aural, although that’d be cool.
- Using visual and aural modes simultaneously is probably a really good idea.
One final thought: I’ve presented this as an aide to my wobbly brain to bring it back up to par, but what is that? Why would you stop at some arbitrary definition of normal? If this tech is any good, it should be able to lift me well beyond what I could manage unassisted, truly augmenting my intellect.
I think technology is pushing in this direction; the smartphones we all use now are a step on the way toward being truly augmented humans. A lot of the things I’ve described above will probably emerge on their own in the fullness of time.
So a lot of the pieces are nothing new, it’s just thinking clearly about why I might assemble them in a certain way.