Science and Technology links (July 21st, 2018)

  1. It seems that something called “Cartesian Genetic Programming” could be a match for Deep Learning.
  2. If you look at the best paid individuals in most fields, most of them are men. Why is that? There are many possible explanations. One of them is that men take more risks. If you hope to ever get larger rewards, it helps to take large risks.

    It is undeniable, I think, that boys take more physical risks than girls. But do men really take more risks than women in their careers?

    Biologists tend to tell us that females are more risk adverse. The usual story is that a female’s ability to reproduce is mostly limited by the availability of food and shelter, whereas a male’s ability to reproduce has far more to do with its ability to impress females, or to impose its will. We can test it out by entering animals in games. For example, Orsini et al. investigated the issue in rats:

    Consistent with findings in human subjects, females were more risk averse, choosing the large, risky reward significantly less than males. This effect was not due to differences in shock reactivity or body weight, and risk-taking in females was not modulated by estrous phase. Systemic amphetamine administration decreased risk-taking in both males and females; however, females exhibited greater sensitivity to amphetamine, suggesting that dopaminergic signaling may partially account for sex differences in risk-taking.

    Social scientists have another, different take. For example, Nelson, an economist, finds the whole question metaphysical:

    The statement “women are more risk averse than men” is fundamentally a metaphysical assertion about unobservable essences or characteristics, and therefore cannot be empirically proven or disproven. A review of the empirical literature, with attention paid to the misleading nature of generic beliefs and statements, the proper interpretation of statistical results, and the quantitative magnitudes of detectable differences and similarities, sheds doubt on whether statements such as these should have any place in an empirical science that aspires to objectivity. The widespread acceptance of such statements appears to perhaps be rooted more in confirmation bias than in reality.

  3. A honey bee has about one million neurons and a billion synapses. The best smartphones have many more transistors than honey bees have synapses.
  4. Do human beings have “general” intelligence? Kevin Kelly, a well-respected writer, does not think so:

    We like to call our human intelligence “general purpose,” because compared with other kinds of minds we have met, it can solve more types of problems, but as we build more and more synthetic minds we’ll come to realize that human thinking is not general at all. It is only one species of thinking.

  5. Our cells are powered by mitochondria (tiny cells living inside our cells). If your mitochondria cease to function well enough, your cells are as good as dead. But could they be revived? It seems so, at least as far as the heart of babies is concerned: fresh mitochondria can revive flagging cells and enable them to quickly recover.
  6. Each time your cells divide, the end of your chromosomes (the telomeres) shortens. Thus a given human cell can only divide so many time. This sounds terrible, but it is less terrible because cells have ways to elongated their telomeres when needed (telomerase)… however, most cells will just divide a fixed number of times. At that point, the cell either dies (apoptosis) or, rarely, becomes senescent. However, the cells don’t literally run out of telomeres. So what happens? According to Ly et al. the problem is structural: the telomeres loop around, hiding the end of the chromosome. When the telomeres are too short to effectively hide the end of the chromosome, it is treated as defective by our biochemistry.
  7. Will we continue to get faster computers? Denning and Lewis think so:

    These analyses show that the conditions exist at all three levels of the computing ecosystem to sustain exponential growth. They support the optimism of many engineers that many additional years of exponential growth are likely. Moore’s Law was sustained for five decades. Exponential growth is likely to be sustained for many more.

  8. There is a huge garbage patch in the Pacific ocean. It is huge: about the size of the state of Texas. A large fraction, nearly half of it, is made of fishing gear.
  9. We making progress against cancer, but it is not enough. In 2018, an estimated 1,735,350 new cases of cancer will be diagnosed in the United States and 609,640 people will die from the disease.
  10. Bill Gates wants to help diagnose Alzheimer’s early. It is believed that by the time you have symptoms of Alzheimer’s, you have already extensive damages. Many researchers believe it could be far easier to prevent these damages than to reverse them.
  11. It is believed that Alzheimer’s might be caused by protein aggregation in the brain. But it is not entirely clear how these proteins causes problems. New research suggests that their proteins induce cellular senescence in the brain. So cell senescence could be a key factor in the emergence of Alzheimers’.

Published by

Daniel Lemire

A computer science professor at the Université du Québec (TELUQ).

8 thoughts on “Science and Technology links (July 21st, 2018)”

  1. Daniel, another reason why there are more men than women among the very successful is that men have higher Std. Dev. of IQ distribution. This would make sense if intelligence is a highly polygenic trait, and if some of the relevant genes fall on the X chromosome: since women have two of those, their influence would average out.

    I think this difference between male and female std. dev. has been known for a long time (I myself have calculated expected gender ratio for Mensa (the high-IQ society) members based on std. dev. difference, and got a number very close to actual – twice as many men as women : -( I wasn’t going to those meetings for the food…). That was more than 20 years ago…).

    It’s part of popular culture too. Just look at the classes for children who in earlier decades used to be labeled ‘retarded’: what do you think, are there more girls or boys among those kids? Assuming the male and female IQ mean is the same, guess what looking at the ‘retarded’ kids can tell you about the other IQ extreme…

    I don’t want even to talk about differences in ambition, and how much career success means to an average men. This comment is already too long…

  2. The statement “women are more risk averse than men” is fundamentally a metaphysical assertion

    This CRETINOUS and it would also be for any other “assertion”: Once you can define a testing protocol with respect to some observable characteristics it may be inconclusive or irrelevant to this or that subject matter but it is not “metaphysical” anymore.

  3. Exponential growth always seems a bit weird to me be cause we live in a resources limited finite world. In chip manufacturing, costs (researching new processes cost as well as lower yields as chips gets more complex) and energy consumption/dissipation are the main brakes to development. Exponential growth cannot be sustained unless we change paradigm along the way. For instance, we stopped doubling clock speed a while ago in favor of multiplying cores. And now, we are slowing down adding new cores to chips in favor of adding more cpus to distributed computing. For me, these are completely different charts and do not share the same Moore’s law line.

  4. Exponential growth always seems a bit weird to me be cause we live in a resources limited finite world. In chip manufacturing, costs (researching new processes cost as well as lower yields as chips gets more complex) and energy consumption/dissipation are the main brakes to development. Exponential growth cannot be sustained unless we change paradigm along the way.

    Moore’s law as originally stated was clearly time limited. However, even the diehards like Kurzweil would agree with that. An idea like Dennard scaling runs its course. Then something else comes along. That is, we are in an open-ended model of innovation. We don’t get one innovation that we have to milk forever. We get a stream of innovations.

    Exponential growth is an empirical observation that goes from way back… from the industrial revolution if not before. We know that it goes all the way to the beginning of life. If you look at the emergence of species, you get exponential curves everywhere. It took a really long time before you had unicellular organisms, then less time to get multicellular organisms then much less time before you got mammals, then much, much less time before you got human beings. If you look at biological time scales, the exponential curve is clear. In technology, you see the same. It took forever to get electricity. Then it took so many years to get the transistors… Progress is clearly accelerating.

    We don’t know whether progress is unavoidable. I think that people like Kevin Kelly expects that it might be. That is, technology is its own thing, and it basically on its own quest, whether we like it or not.

    What is clear also is that the future is super hard to predict. It does not follow nice trajectories. That’s pretty much what “open-ended” means.

    For instance, we stopped doubling clock speed a while ago in favor of multiplying cores. And now, we are slowing down adding new cores to chips in favor of adding more cpus to distributed computing. For me, these are completely different charts and do not share the same Moore’s law line.

    Right. But I think that’s somewhat about semantics. We are making progress… maybe not exactly how Moore’s imagined it, but Moore never thought about 2018 when he drafted his law.

    1. It’s interesting that you mention biological evolution. My point about limited resources particularly applies to the human population. We saw exponential growth in the number of human beings but that trend cannot continue very far in the future. Last time I checked, earth could not sustain more than 10-15 billions peoples. Same thing for CO2 emissions. We are bound to stop exponential growth sometime in the near future otherwise we are doomed.

      Going back to technological progress.. I can’t tell where we are in the curve but it got to slow down eventually. We face limitations in terms of economical costs to develop. My view of the world fits more an array of S curves than a single infinite exponential curve 😉

      And yes, that was only semantic interpretation of the topic..

      1. Earth’s population is not growing exponentially. Everywhere but in Africa, we are aiming toward a near-term plateau, followed by a fall. Germany and Japan are already in massive decline. Simply put, educated women rarely decide to have many children early on.

        This population decline is happening just as most of us have incredibly long life expectations. The two appear to be related: the longer we live, the slower we reproduce.

        A natural extrapolation is one where few of us reproduce but we all live much longer.

        This being said, how many people could make up humanity? I don’t think we know, and I think it may be far greater than anyone might estimate currently. Certainly, if your refer to primitive human beings, the answer is relatively small… but if you refer to technological human beings (and all of us are, one way or another, technological human beings), then the question is how much technology are we talking about?

        Prior estimates of how many human beings could live on Earth have been far exceeded without harm. In fact, fewer people starve today and obesity is a problem.

        Why were the prior estimates wrong? Because they assumed constant technology…

        We live in larger houses than our ancestors or even than our grand-parents. More of us have private bedrooms, private bathrooms, and so forth. We have lots more food.

        That is, there are more of us, but we also have more food and more room!!!

        You have to take into account that each additional individual is more than just a mouth to feed, it is also a potential for invention, new science, new engineering and so forth. The more of us there are, the more quickly we can innovate and the more complex a civilization we can sustain.

        Run an experiment. How much complexity could you and I on a deserted island sustain? Surely nothing like our current technology. We’d be lucky to survive a few years.

        However, with billions of us, we can make incredible progress. It is enough for one of us to solve a hard problem for all of us to benefit years later. Thus, having more people is not necessarily a drain. It might be, in fact, critical for our sustainability. It is not at all clear that it is a good thing to have fewer people. Fewer people means less innovation and we need constant innovation to be better off. Thankfully, as the number of human beings is set to start decreasing during this century, we are also in the midst of a cognitive revolution (e.g., computers).

        Why would technological progress slow and come to halt? I can tell you that if it does, it is very bad news. It is not at all clear how we make people better off without technological progress.

        Long-term stagnation did happen in the past. The fall of Rome was not great for technology. But it was also not good for the people.

        If you are worried about things like energy use… you should be aware that we are generally using less energy today than we did in the recent past… We are hitting a plateau. Look around you. I am typing this on a tablet that uses a fraction of the energy that my PC used 20 years ago.

        Technology does not necessarily require more ressources.

        Anyhow, there is no way to be certain about the future. Certainly, we can hit a plateau. My main point here is that if it should happen, it is not going to be good for the people.

          1. Even so, try Africa where the population is still increasing exponentially (at least in large parts of Africa) and look at the food consumption per capita. It is going up! That’s not what a model indicating that we are about to run out of ressources would predict.

            (Of course, people generally eat too much and we should lower the calories per capita. But that’s another story.)

            Look also at the forest cover: it is going up. That’s right… in most rich countries (Europe, North America) and several not so rich countries (e.g., China or Russia), there are more forests today than before.

            You read this right: more people and, yet, more forests. Part of the trick is that human beings actually occupy less land than we used to!

            Where do you live, Michel? Probably in a densely populated area, right? What about your ancestors? Probably in a not-so-densely populated area, right?

            So as time passes, we aggregate closer and closer together.

            That sounds terrible… less space per person, right? Except no… with technology, we create space.

Leave a Reply

Your email address will not be published. Required fields are marked *

To create code blocks or other preformatted text, indent by four spaces:

    This will be displayed in a monospaced font. The first four 
    spaces will be stripped off, but all other whitespace
    will be preserved.
    
    Markdown is turned off in code blocks:
     [This is not a link](http://example.com)

To create not a block, but an inline code span, use backticks:

Here is some inline `code`.

For more help see http://daringfireball.net/projects/markdown/syntax