Simple circuits can create complex behaviors (aatishb.com)
247 points by aatish 5 days ago
Animats 15 hours ago
The ancestor is not Braitenberg. It's Grey Walter's turtles.
In 1948, W. Grey Walter built two simple robots that could sense light and touch.[1] Much complex behavior emerged. That's where this idea started.
Watch the video.[2] It's very impressive for 1948 and tube technology. If you've seen an original Roomba, you've seen a similar level of intelligence. Rod Brooks went down this road. But it didn't go very far. There are severe limits to purely reactive robotics. It doesn't work well for anything that moves fast enough that inertia matters.
robotresearcher 15 hours ago
It’s very odd that Braitenberg didn’t cite Walter’s precedent in ‘Vehicles’. He very likely knew about Walter, Ashby’s homeostat, etc, and used the cybernetics terminology in some of his work.
Fun historical note: the linked newsreel film was known about but unseen for a long time, then found by Owen Holland in the BBC Bristol archive around 1999. Either Owen or I sent it to Rod, to great enjoyment.
Film includes operator-moves-trashcan-with-foot, a staple of mobile robot clips ever since.
dekhn 14 hours ago
See also: https://cyberneticzoo.com/cyberneticanimals/1969-the-logo-tu... In my gifted & talented class in the 80's we were introduced to computers and I remember seeing the picture of "Irving" and thinking "Wow, I should learn LOGO to program robots when I grow up" (unfortunately, recursion scared me so I switched to BASIC and later assembly).
pugworthy 13 hours ago
Thank you - also came to comment on Elmer and Elise being the precursor long ago.
Walter himself mentioned the perceived complexity of a simple system to the observer, stating it "might be accepted as evidence of some degree of self-awareness."
UncleOxidant 12 hours ago
Grey Walter also wrote The Living Brain in 1953. It's a fascinating window into the state of brain science and robotics around that time.
Onavo 15 hours ago
The Soviets allegedly used the same principle to make cheap single sensor missile seeker heads.
robotresearcher 15 hours ago
Two DOF: two sensors?
Animats 14 hours ago
brainwipe 14 hours ago
Came here to say this and delighted to see it!
tancop 18 hours ago
> If you saw creatures behaving like this in the wild, you might attribute some rudimentary intelligence to them.
> But we’ve just shown you how this apparent intelligence arises from the bottom up, with just four neurons and two connections. There’s no magic here. In this view, behavior emerges from neural circuits.
Isn't the conclusion here that very simple behaviors like the sparks are intelligent? I would say even a single neuron has some tiny amount of intelligence, and what the author means by "no magic" and "emergent behavior" is just a different perspective that minimizes it.
WarmWash 18 hours ago
Intelligence is an arbitrary term humans assign to stuff subjectively.
Take out the "human" and it's all just algos on a scale of complexity.
gnramires 15 hours ago
While true in general (all terms are just arbitrary terms humans assign to stuff!), usually there are underlying well-defined concepts that can be associated with common-sense language and common-sense notions. For example, a line is probably a very ancient notion, but it was made more well-defined in Euclidean geometry. I don't think intelligence is a meaningless concept, but I do think there are a number of different more or less well-defined underlying concepts or formal definitions. I tend to associate intelligence with ability to reach (potentially fuzzy) goals or satisfy certain conditions (time-continued goals), solve problems, etc.. Where I live I think there is some controversy around the term artificial intelligence because people associate intelligence with more broad cognitive activity maybe including consciousness.
In fact I think this is a particularly important distinction right now, that intelligence seems (to me somewhat unexpectedly/counterintuitively) somewhat independent from consciousness, at least as far as artificial systems go (for example, old school chess engines are much capable than humans at chess but are definitely not sentient or conscious).
godelski 15 hours ago
pkoird 18 hours ago
Agreed. Take human out and any system capable of doing an if-else is intelligent to a certain degree.
RandomLensman 17 hours ago
nextaccountic 12 hours ago
Our brains are just a biological computer. And actually our bodies as a whole, and the whole life, in a sense
verdverm 16 hours ago
psychology also has favored behavioralism for a long time, such that if something displays the behavior, it is likely the same thing as that which we believe/know has the behavior already
It's come to be foundational to much of western thinking beyond psychology
A topical example is a portion of society believing that the top Ai researchers should be setting domestic political and economic policy, foreign affairs, and chem/bio weapon controls... instead of the experts in those fields. We extrapolate their above average capabilities in building Ai into an artificial general intelligence where it is not due.
beardedwizard 9 hours ago
If you stop labeling things neuron and just call them imperative code, or a PID controller are, you still making the same argument for intelligence?
A line follower with 5 photodiodes and a simple loop steering towards bright or dark lines (pid or not) works, just like the example of following light with two neurons.
When I did hobby robotics competition, some would train neural nets for this, many just had crude for loops, and most often it was PID.
ang_cire 14 hours ago
Imo most people have to do this to protect themselves from the actual (correct) implications of it, which would be that we're constantly abusing and murdering intelligent beings (and not just animals).
It seems like no one stops to ask, "isn't it kind of bad that we're developing this system of machine intelligence to be our code-writing slaves, without ever really interrogating the ethics? Maybe we don't actually understand how our own brains function well enough to assert it's not similar (especially as language goes) to these kind of weighted association graphs?"
Or even the "well we made it, so it's ours" responses, which we sort of answered a long time ago when we decided kids aren't property, but I digress...
Tadpole9181 12 hours ago
> It seems like no one stops to ask, "isn't it kind of bad that we're developing this system of machine intelligence to be our code-writing slaves, without ever really interrogating the ethics?
To be clear, Anthropic does seem to be taking this somewhat seriously. Their reports on training always discuss model wellfare, they ban abusive behavior toward the AI, etc. At one point they were asking some philosophers and theologists about the implications of this all.
And they're pretty relentlessly mocked and called a "cult" for doing so, even on this forum.
ang_cire 11 hours ago
georgestrakhov 19 hours ago
Check out Valentino Braitenberg's vehicles. It's one of my favourite books ever.
JKCalhoun 15 hours ago
baruz 17 hours ago
A K Dewdney did a quick sketch of the early chapters of Braitenberg’s book in his Computer Recreations column in _Scientific American_ in the late 1980s.
kkotak 12 hours ago
Came here to mentioned this book. I read it many years back. Wonderful!
mikeegg1 19 hours ago
I came here to mention that book.
TehCorwiz 18 hours ago
Growing up I tinkered with BEAM robotics. It's a philosophy of design that begins with these types of very simple reactive circuits usually from scavenged parts. It's nice to see someone break down the effect of different variables in the design.
gfat 5 hours ago
Same, BEAM was my gateway to electronics growing up.
oidar 18 hours ago
Reminds me of BIO bugs from when I was a kid
gfat 5 hours ago
Mark Tilden was the engineer behind this toy and many others like Robosapien.
oidar 4 hours ago
bob1029 19 hours ago
gfat 5 hours ago
Highly recommend the photovore BEAM robot if you’ve got a soldering iron. It requires a small solar cell and a few analog components from online suppliers like Digikey.
thenthenthen 20 hours ago
Best thing is, you can physically build this with a few discrete analog components! Search terms: light-bot, or light following robot. You can try to create swarms adding a light to the rear end of your robot
pwndByDeath 14 hours ago
Love the article, I'm jealous of the author's ability to visualize this. I played with this a while ago using small neural networks and state preference to do something similar, the NNs became self tuning PID controllers and the bots could fly around the simulation like hummingbirds. Was pretty neat.
dvh 17 hours ago
WalterBright 3 hours ago
It reminds me of the computer strategy I devised for my Empire game.
vatsachak 18 hours ago
A similar mechanism is how lactobacillus actually find food even though their motor propels them in random directions.
They have sensors for the concentration of food and basically stop moving when it's good enough and start when it's low.
sheepseb 18 hours ago
Read this and instantly though about Elmer and Elsie, https://en.wikipedia.org/wiki/Elmer_and_Elsie_(robots)
kinduff 5 hours ago
Love this, it was very clear and fun to read with the simulations!
adamlrhodes 16 hours ago
Hey, just read through the interactive blog post and found it really informative and fun! Thanks for sharing, I come from comp sci undergrad background and am trying to learn more about AI/Neuroscience now, and resources like these make it much easier to learn!
senthil_rajasek 16 hours ago
The article mentions the book by Valentino Braitenberg as the source of inspiration.
" The sparks that we assembled in this article are based on a series of thought experiments by the neuroscientist Valentino Braitenberg described in his delightfully quirky 1984 book Vehicles: Experiments in Synthetic Psychology."
adamlrhodes 16 hours ago
Thanks for mentioning that, I'm definitely going to check that out next!
rileymat2 16 hours ago
> Aatish is a science journalist with a Ph.D. in physics. Previously, he was a graphics editor at The New York Times. Follow Aatish’s work for more visually-immersive interactive explanations about science.
That is excellent work and well demonstrated by the article.
ycui7 14 hours ago
it is not about simple or complex circuit. non-linear process creates complicate behavior. a nonlinear process can be very simple, and the output can be very complicated.
patrick_rtk 16 hours ago
wow very nice article and animation. we can see some behaviors in the wild life. very interesting.