Hello,
New to the forum. I’m not highly technical, rather I work alongside highly technical people designing and building both physical and digital tools that are easy for humans to use. My responsibility is the usability part.
I’d like to use 1KBP as a tool to improve application design, how quickly and easily a human can use a piece of software to perform tasks.
I don’t expect 1KBP to do any actual design, more along the lines of feeding it a base design and process flow and see how long it takes to “learn” it from beginning to end. Feeding the system a few initial designs for the same task could get the product further ahead for a better out-of-the-box experience for a typical user.
Extending further, once a decent baseline is chosen, IKBP could be used to identify specific areas within the design that are barriers and take the most time for it to get through the task. Design iterations for process order, design and feedback mechanisms can be done to optimize for time saved. When new features are considered 1KBP could be used to simulate usability impact on an existing user base.
My objective is to make tools that save the most time and require the least effort on the brain to use. At least from my very high level view, it seems like 1KBP could be exceptionally useful in that regard. If I’m off base please let me know. Thanks.
Hi there, the Thousand (1K) Brains Project is not capable yet of learning “tasks”, especially involving 2D software UI. Right now, it’s a prototype geared toward 3D object analysis. The team is hard at work figuring out how to tackle behavior learning, which may one day evolve into long-horizon task learning, but I think it’s a little too early for TBP to be used the way you’re looking for.
If anything, I don’t think there’s any type of AI that would currently be reliable for this. You still need experience-driven human feedback to tell what works best for other humans 
However, since TBP aims to replicate how the human brain works, it might one day become suitable for this sort of application, in the long term.
(Disclaimer: I am not part of TBP staff, just a community member)
2 Likes
Welcome aboard!
As @AgentRev notes, the TBP is still very much an early-stage research project. However, (IMHO, as an interested observer) that doesn’t mean it’s silly to consider and discuss usability, etc. For example, as Monty implementations start to scale into lots of modules, there will be some real challenges for data presentation, graph visualization, etc.
That said, while searching for both “1KBP” and “IKBP”, I found nothing that seemed to be relevant. Could you provide a link or two and perhaps some notes on what this might bring to the party?
2 Likes
Hello,
Thanks for the info. The difference as I see it is 1KBP may eventually end up with some of that functionality even if it’s a long shot, the vast majority of the other platforms won’t ever get there. I see simulating even a incredibly small part of the brain in a very rough and primitive way as a beneficial tool for this sort of application. Being able to remove experience for the equation could remove personal bias from initial designs and give the designer an idea how easy their design is to pick up for the user. Give the person another data point to consider in the design, not design for them. I don’t believe the pattern matching stuff will ever get close to being an assistive tool in this way, garbage in garbage out. Make better assistive tools to give the person more time to concentrate on the actual task rather than load them up with tools that steal brain power and waste time.
Yep, 1KBP = TBP
