"Using LEGO robots to study posthumanism at the University of Alberta. Part 1: Parable of the ant

Music: Phillip Glass - Metamorphosis One"

https://www.youtube.com/watch?v=MtNZYIeklds&list=PLADA9C21F511A7BF6 https://mastodon.social/media/35csIf0o5kofaCohlmk

"Walter's tortoises represent the first real world demonstration of artificial life."
https://www.youtube.com/watch?v=lLULRlmXkKo

Just some kid mesmerized by "B.E.A.M. robots" trying to recreate living things with a lego set.

I latched onto the concept of environment as feedback loop. I independantly recreated the concepts behind Walter's tortoise and Braitenberg's vehicles.

While schools cargo-culted ideas behind line-following robots and fetishing turtles, I was studying neurons.

Walter was a neurologist.

"Whats happening here? Is it something electronic? Something Mechanic? Are we modifying the world? Maybe it's a combination of all of them..."
https://www.youtube.com/watch?v=7ncDPoa_n-8

"proctoring a silicon species into sentience, but with full control over the specs. Not plant. Not animal. Something else."

"I see them as the components of a programmable ecology. They'll replant forests, hunt cockroaches, monitor poachers, cut your grass, clean your pool, polish your floors - all invisibly, dependably, for years."

https://www.wired.com/1994/09/tilden/

Computational Metapsychology
https://media.ccc.de/v/32c3-7483-computational_meta-psychology

"This is why I talk to myself using social media.

It lets past me communicate with future me."
https://mastodon.social/@ultimape/14359586

"The idea of the Turing test is that you build a computer that can convince its audience that it is intelligent. But when we look for intelligence, we actually look for something else: we are looking for systems that are performing a Turing test on us."
https://www.youtube.com/watch?v=3Bns4HkAniA

*Looks around, Sees how impressed everyone is...*

Actually, I even doubt my own intelligence and perform the Turing test on myself.

I seem to pass, but results are inconclusive: I may or may not be a p-zombie with false memories.

Combine Braitenberg's vehicles with B.E.A.M. robots and try to recreate a minimal set of swarm oriented primitives in a cellular like manifold?

Yes please. https://www.scientificamerican.com/article/ldquo-smarticle-rdquo-robot-swarms-turn-random-behavior-into-collective-intelligence/

I have yet to see an object detection algorithm that works on video that isn't simply a single image detection pass. Even the latest on TED https://www.youtube.com/watch?v=Cgxsv1riJhI is just a single image detection system running in real time.

Is anyone working on Object Permanence models of machine cognition?
https://en.wikipedia.org/wiki/Object_permanence

I've also not seen anything using Optical Flow that even approaches what we know about human vision systems approach it. This thing with bees is close:
https://www.youtube.com/watch?v=olZY7yBD0w4

An object permanence model would solve issues like "based on the past detection, I am 90% confident I am inside therefore if I see a door I shouldn't tag the image as "outside"... or "hey, that person is wearing a stopsign on his t-shirt, this isn't an actual stop sign"

The stuff they are doing at skydio has the potentially to move in that direction: https://www.youtube.com/watch?v=P3E4pl2Weos

Esp. if they combine their tech with https://www.youtube.com/watch?v=H7Ym3DMSGms

Can someone pay me to build swarming robots. I have been thinking about swarming robots since I was 11.

Collaborative ratcheting Bayesian sensory networks. https://www.cambridge.org/core/journals/psychological-medicine/article/what-is-mood-a-computational-perspective/5FA0177A965FF3EE01D4AA5C09C0A2A5

One of my most detailed memories of my childhood was a game I'd play where I'd imagine what it was like to 'see' as a hive mind like creature - multiple views and bodies with different thoughts, yet all collaborating. Thunder Cats and Power Rangers, but I was the zoid.

I would spend hours just imagining things - worlds, people, animals - then building them out of legos and lincoln logs. Entertaining myself, alone.

All of this was before I had access to a TV beyond PBS.
https://www.youtube.com/watch?v=OFzXaFbxDcM

I realized that most of our machine learning models for emotion detection are using still frames. You can freeze a frame of someone talking and the computer will detect that they are happy, angry, or other glitched emotions because of the way the face moves for certain words.

Training a machine to make you smile might only make you talk.

Pathological cases: say cheese over and over or else it gets the hose again.

A different issue with machine learning of face=>emotion is that many faces used in databases are biased toward emotions from white anglio-saxon males.
Most don't consider differences between fake smiles either.

Do any amount of research on cultural expressions of emotion and you'll find many of them are learned behaviors. I say this as a face-blind autistic who's trying to learn to read emotion myself.

Stuff like this https://www.forbes.com/sites/johnnosta/2018/01/04/google-glass-is-a-hit-for-children-with-autism/ "works" but it is limited to particular cultures.

If you dig into the source of many of these face datasets: they often don't include people with facial abnormalities or emotive variation. There's all sorts of disorders that lead to physical issues with the muscles of the face. Stuff like Moebius Syndrome, or other muscle issues.

You can test this yourself. Take stills someone talking on YouTube and run it thru Microsoft's face API: https://azure.microsoft.com/en-us/services/cognitive-services/face/

I often smile with only my eyes. The forced grin thing my culture does is lost on me.

Smiles themselves are fascinating from a psych perspective: http://www.businessinsider.com/a-neurologist-explains-how-to-spot-a-fake-smile-2016-7

But there is also a layer of communication and game theory involved if you really wanna dig in: http://users.ox.ac.uk/~kgroup/publications/pdf/scharlemann_etal_2001_jeconpsy_smile.pdf

Part of it might explain why American's have a weird smile that creeps out tourists. https://www.youtube.com/watch?v=ojbJrdkPhGg

All sorts of weirdness seems to pop up around how cultures read emotions too. The bias in autism's attention on mouths https://academic.oup.com/scan/article/1/3/194/2362950 has been particularly illuminating to me.

This isn't to say i'm against advancements in machine learning emotion detection. I am just not as excited about the current methods helping me overcome my issues.

"we urgently need tools to help individuals who have both autism and alexithymia understand their own and other people’s emotions."
https://spectrumnews.org/opinion/viewpoint/people-with-autism-can-read-emotions-feel-empathy/

I *am* afraid of what people are doing with these shit datasets. Do they think I'm angry all the time because I don't smile? I get that enough from humans, thank you very much.