Friday, March 23, 2018

Gut Buster


Reading Two Books, by William Wegman, 1971

The Limits of Explainability - Why Artificial Intelligence Needs To Learn How To Follow Its Gut
Mar 2018, Joi Ito for WIRED

Researchers at MIT led by Josh Tenenbaum hypothesize that our brains have what you might call an intuitive physics engine, a noisy Newtonian system. And that system will need to be better understood if we are to make the robots more like us. (Is that the point? I think sometimes it is, yes.)

>>> 

After reading this article, I'm not really sure what's meant by "giving AI its gut" (maybe it's because I've been listening to so much panel discussions on AI from people who are really specific about these things?).

I will assume that to give AI its gut means to give it the ability to solve problems intuitively. The same argument I would use here to express the difficulty in achieving this is the one used to show that we can never create a robot that can smell.

To smell, by the full definition of what it means to smell something as a human, requires a lifetime of experience (at least a childhood's worth). We do not "know" that shit smells bad; we are taught, either actively or passively. And every smell we recognize is done so against a massive memoryplex accumulated and constructed over our lifetime.

We also do not understand the way our olfactory circuitry works, not well enough to reverse engineer it into a robot. For each person, the circuits work differently, or at least differently enough that we have no model for it.

Intuition is difficult to put into an algorithm for similar reasons. It is distilled from experience, and the rich, multi-sensory, emotionally-laden, and socially-mediated experience that comes from human existence. To write a line of code for every 'experience' is ridiculous. But how do you get intuition without the experience? That's going to be a hard problem to solve, and there are many other problems that will need to be addressed before we get there.

The hardwiring problem is similar between olfaction and intuition in that we don't know how intuition works in the brain. We don't know 'where' consciousness is in the brain. We don't get the models so how could we design such a system?

I'm not saying this is the solution, just that it seems to be the obvious extension of the endeavor - to get robots that do all the things that we do, we need those robots to first be born and to live. In a simulated world? Sure, maybe, if that comes first.

Post Script

On the topic of Artificial Intelligence, check out a couple of these recent discussions:

Parliamentary Debate on the Future of Artificial Intelligence
Hosted by Steven Pinker, 2017

2018 Isaac Asimov Memorial Debate - Artificial Intelligence
Hosted by Neil DeGrasse Tyson, 2018



Thursday, March 15, 2018

Genes, Skin Creams and You


Jan 2018, Wired

In short - BabyGlimpse is a $259 test using parents' DNA to predict not only physical features such as hair and eye color, but also taste preferences. (It's true, there's genes that make sugar "taste" better to some, broccoli taste worse, and cilantro taste like soap.) This is called direct-to-consumer genetic testing, and it's a budding industry.

It's supposed to be for fun, but public health experts aren't laughing.

At least that's the angle of the WIRED article I'm looking at here. They also bring up the problem with labeling people. For example, there's a gene that codes for your ability to recognize noise patterns, and could make you a musician, or not.

Genes are complicated (I believe the correct phrasing is "genomics are complicated"). Environment and epigenetics change the way your genes are expressed. Whether or not you eat your vegetables, whether you're skin is one color vs another in a country that was built on such a distinction, those are "environment". Whether your mom was pregnant during the Great Potato Famine, and whether, again, your mom grew up subject to the stresses of racism, those are epigenetics.

You can't predict how those things will affect your genes, and so to label someone good or bad at something at birth could effect their outcome in life. A scientist consulted for the article calls these results the genetic equivalent of skin cream for wrinkles (which is a pretty good dis if you ask me.)

All we really care about here at Limbic Signal though, are the hard stuff - the genes that make your hair brown or your eyes blue, or your armpits smell or not.

Anosmia, or smell-blindness, is caused by a defective, or inactive gene. We have lots and lots of genes for smell, more than any other gene family, in fact. And we are still evolving in regards to smell. And look at how dolphins, who I guess used to be land-dwelling mammals, have had most (all?) of their smelling genes turned off now that they live in the water, and (I guess again) they will have to continue to live in the water for a very long time until they develop again, if they ever do.

We humans are still in the process of selecting which things we should be able to smell and which things to not waste our time on. And in the meantime, we differ greatly in the genes which have been turned off. Some folks can't smell semen, some can't smell rotten fish, some can't smell farts. For the most part, these folks won't be at a disadvantage as a result of their smell-blindness (unless, of course, in the case of not being able to smell a natural gas leak, but that's a big exception.)

Nonetheless, if you find exciting the prospect of knowing in advance if your child will have a specific anosmia, then this "wrinkle cream" is for you.

Post Script

Some other cultural evolution things, from sister blog Network Address:




Tuesday, March 13, 2018

Domestic Chemical Signature


1950 Commercial for Lysol


Feb 2018, phys.org

It comes as a surprise to me sometimes when I read that “no one really knows” about the most basic things. For example, I see a headline that says – Science Finally Understands Why Ribbons Curl (you know when you the back of a pair of scissors on decorative packaging ribbon and it curls?). How do we have a roaming robot on Mars and not know about things like this?

In the article above, science wonders what kind of chemical reactions are going on inside our homes.
The stuff we use to clean our homes, our clothes, and ourselves, as well as the stuff that gets in the air when we cook, as well as the general environmental pollutants around us, these are all chemicals that interact with each other, and we don’t know the full extent of these reactions.

Cigarette smoke reacts with gaseous nitrous acid in our homes (a waste product from cooking and heating, and a general environmental pollutant from car exhaust etc. ) which is then absorbed by wood furniture and the like. So yes, there is a particular smell for a house that’s been smoked in for twenty years vs. smoke itself.

Here’s another one – chlorine bleach, as a gas in the air, oxidizes stuff, which in combination with UV light forms radical molecules.

You don’t have to be a chemist to know that everyone’s home smells different. And that difference is made by all the chemicals we use in our daily lives. The thing is, if you were to try and pinpoint the source of that smell – the chemicals – you would fail. Chemistry is a thing in flux, it is always changing, and even Plato recognized it as such when he described odors in his Timaeus as “emanating from a body undergoing change.”  Therefore, the odor fingerprint of your home comes not only from the myriad substances you use in your house, but from the reactions of those things with each other.

And if Science doesn’t even know how all that stuff works, then how are you supposed to know?


On a side note:

Cleaning products as bad for lungs as smoking 20 cigarettes a day:
Scientists advise that harmful sprays could be replaced with simple microfibre cloths and water
Feb 2018, The Independent

And on that note:

This one is about cleanliness and the illusion of cleanliness. Lived in a basement apartment for awhile and decided to pay a friend of my landlord to do a “deep clean”. She brings her own cleaning products (because I use mostly microfiber and water, and lots of people either don’t know how to use that or don’t believe that it works.)

So this lady comes in and cleans up my place. Hella Lysol. My landlord comes down to visit that day, to see what she did, and boy is she delighted. “It’s so clean in here, wow you really needed it!”

First of all, my apartment is not dirty. I was a custodian at almost every job I had, not to mention keeping a high school art classroom in pristine condition for ten years, not to mention that this very apartment had been hit by a massive mold attack one summer, and I’ve been hypervigilant/paranoid about its return ever since. And, not only am I the guy that cleans for the maid, but this time at least, I had to clean again after the maid left.

My landlord, however, is suffering from olfactory insinuation. She smells Lysol and thinks it’s clean, like, cleaner than it was before. Funny, because as soon as I walked into my apartment, you know what I smelled? “Dirty mop” that’s what I call it.

“Dirty mop” is when you don’t clean the mop itself after mopping, and you also don’t let it dry, and you also keep it in a dark place until the next time you use it. And when you do use it next, you, instead of cleaning, spread that dirty plethora of microbes all over your floor.

So I smiled, paid the woman, and got to work cleaning my apartment, and wondering how long it would take for this Lysol smell to dissipate.

Thursday, March 8, 2018

Primitive Mentality


Conservative caveman. Just kidding, he isn’t actually disgusted, because his lip doesn’t curl back to resist the (simulation of) bad food that would be approaching his mouth.


Mar 2018, phys.org

“People who are easily disgusted by body odours are also drawn to authoritarian political leaders.”

That’s what the article says. Well, that’s what the scientist hypothesizes. And his explanation is pretty convincing – if you’re extra sensitive to the smells of sweat or urine, you may be more supportive of the dictator type of political leader who “puts people in their places,” because that type of leader reduces contact between different groups of people (i.e., divides people) which decreases anyone’s chance of becoming ill.

Say that again? The further around you turn your head when you smell [other people’s smells] the more you appreciate a political leader that makes sure you don’t have to smell “other” people.

Usually, I take smell-in-the-news with a grain of salt. Reason being, smell is too misunderstood of a sense to make groundbreaking discoveries in psychology by way of its perceptual channel. Usually also, I take behavioral /social psychology in the news with another grain of salt. The reason for that is the replication crisis in psychology (particularly social psychology, similar to behavioral psychology).

That being said, Jonas Olofsson of Stockholm University has an interesting idea. Disgust is a powerful emotion, and one that can, or should, be easily tested via smell. A little etymology digression – disgust comes from gustation which is about taste and eating and satiation, and to figure out what’s good to eat is one of the few main reasons why we smell, or so it seems. That natural behavior of determining what’s good to eat extrapolates into all kinds of other behaviors, one of them possibly being who to vote for.

Psychological theories about disgust and (American) conservativism have been around for quite some time, but this is a different approach. Check it out for yourself and see what you think.

Thursday, March 1, 2018

Hunting, Gathering and Gene-making


The Visionary Origin of Language, by Alex Grey
Jan 2018, phys.org

In Asifa Majid news:

There is only one name you hear on the subject of the language of smell while reading your weekend science news, and that's Asifa Majid.

She studies the indigenous people of the Malay Peninsula of Southeast Asia, a people who she has shown to have a remarkable smell-vocabulary. They call this skill "olfactory naming".

In this particular article, Majid speculates as to whether this special vocabulary is natural to hunter-gatherers, or just the particular people who she studies. Turns out that horticulturalist-people who live right among the hunter-gatherers, and who even speak a very similar languages, do not have the same a proficiency for labeling odors.

The hunter-gatherers out-performed their horticultural counterparts. (They are better at naming colors too by the way.) Because everything else is the same between these two groups of people, the findings suggest that culture, rather than hard neuroarchitecture, is responsible for the difference.

Following this line of reasoning, anyone can be "taught" by their culture to be good at naming odors; it's not beyond our abilities as humans; it's not something that has been evolved out of us.

Read her paper here:
Current Biology, Majid and Kruspe: "Hunter-Gatherer Olfaction Is Special"


Admittedly reaching far beyond the scope of this paper, I'll also mention that human evolution as a hard, genetic phenomenon, has been usurped by cultural evolution. A great inflection point is the adaptation of some humans to produce the enzyme lactase into adulthood. This gives them the ability to digest milk as adults, a feature present in no other mammal.

It may be possible that the genetically determined shape of our skulls and jaws, a shape that separates us from our primate ancestors, is also a result of "culture" but we can't prove that, and it's unlikely that the timelines would match up - a culture that would last long enough to change the physical structure of the body.

Milk digestion is most certainly a result of some cultures practicing dairying for centuries (maybe millennia). And so, in effect, this is the first, or most salient, example of culture changing the human genome.

Digesting milk and naming odors are not the same thing, because one is a hardwiring issue, and one is softwiring. This was in part the point of Majid's paper to ascertain.