Templeton Prize winner spouts more nonsense in Scientific American

March 22, 2019 • 10:15 am

The other day, theoretical physicist Marcelo Gleiser of Dartmouth College won the £1 million Templeton Prize for “affirming life’s spiritual dimension.” At the time I used his quotes reported by the media to show that, while Gleiser may be a good physicist, he’s not exactly the sharpest knife in the drawer when it comes to talking about atheism and “the limits of science.”

This is all confirmed in a new Scientific American interview with Gleiser conducted by Lee Billings, an associate editor who writes about cosmology and physics. Read on, but don’t waste a lot of time:

It’s the usual pap espoused by scientists who are also “spiritual” in the way Templeton likes: touting the limitations of science, saying there are other ways to answer the Big Questions, and, especially, dissing atheism.

One of the most disingenuous parts of the interview, and a sure sign of thoughtless accommodationism, is the call for “humility.” Of course, that call applies only to atheists and scientists, not to believers:

[Billings] Right. So which aspect of your work do you think is most relevant to the Templeton Foundation’s spiritual aims?

[Gleiser] Probably my belief in humility. I believe we should take a much humbler approach to knowledge, in the sense that if you look carefully at the way science works, you’ll see that yes, it is wonderful — magnificent! — but it has limits. And we have to understand and respect those limits. And by doing that, by understanding how science advances, science really becomes a deeply spiritual conversation with the mysterious, about all the things we don’t know. So that’s one answer to your question. And that has nothing to do with organized religion, obviously, but it does inform my position against atheism. I consider myself an agnostic.

It always sounds good to tout “humility”, doesn’t it? The thing is, scientists are already humble, because we’re forced to be. Yes, there are arrogant scientists (Lynn Margulis comes to mind), but as far as practicing science goes, you’re always looking over your shoulder asking “What if I’m wrong?” “How can I make sure that there are no flaws in my work that others might see?” You don’t become famous by being loud (although sometimes that helps); you become famous by being right. And the more arrogant you are, the more likely others are to replicate your work.

Do you know this famous quote by Thomas Henry Huxley?

“Sit down before fact like a little child, and be prepared to give up every preconceived notion, follow humbly wherever and to whatever abyss Nature leads or you shall learn nothing.”

Gleiser’s double standard for humility becomes clear when he goes after atheism but doesn’t go after believers:

Why are you against atheism?

I honestly think atheism is inconsistent with the scientific method. What I mean by that is, what is atheism? It’s a statement, a categorical statement that expresses belief in nonbelief. “I don’t believe even though I have no evidence for or against, simply I don’t believe.” Period. It’s a declaration. But in science we don’t really do declarations. We say, “Okay, you can have a hypothesis, you have to have some evidence against or for that.” And so an agnostic would say, look, I have no evidence for God or any kind of god (What god, first of all? The Maori gods, or the Jewish or Christian or Muslim God? Which god is that?) But on the other hand, an agnostic would acknowledge no right to make a final statement about something he or she doesn’t know about. “The absence of evidence is not evidence of absence,” and all that. This positions me very much against all of the “New Atheist” guys—even though I want my message to be respectful of people’s beliefs and reasoning, which might be community-based, or dignity-based, and so on. And I think obviously the Templeton Foundation likes all of this, because this is part of an emerging conversation. It’s not just me; it’s also my colleague the astrophysicist Adam Frank, and a bunch of others, talking more and more about the relation between science and spirituality.

I’ve already discussed this cockeyed view.  The idea that atheism is “belief in nonbelief” is a Deepity, for it sounds profound but upon examination proves shallow and, indeed, stupid. Atheists don’t “believe in nonbelief”: they reject acceptance of gods because there is no evidence for gods. That’s all there is to it, and, contra Gleiser, it’s absolutely consistent with the scientific method. In fact, the refusal to accept “truths” when there’s no evidence for them is the hallmark science—although there is no formal “scientific method.”

In contrast, Gleiser’s own agnosticism is simply a chickenshit way to avoid conclusions that he’d draw in any other area of science. If one posits that there are aliens living on Mars, but we find no evidence of them after repeated scans of the planet, and no signs of life from sending up biological probes, then you’re being unscientific. As Vic Stegner used to say, “The absence of evidence is evidence of absence—if that evidence should be there.”

As I wrote in Faith Versus Fact:

In science, if there should be evidence for a phenomenon but that evidence is consistently missing, one is justified in concluding that the phenomenon doesn’t exist. Examples are the Loch Ness Monster and Bigfoot, as well as paranormal phenomena like ESP or telekinesis. Seeking evidence for such things, the skeptics always come up dry. It is the same with God, though theologians will object to comparing God to Bigfoot. The philosopher Delos McKown had a more parsimonious answer for God’s absence: “The invisible and the non-existent look very much alike.

And I said this after examining Barbara Forrest’s convincing argument that the success of methodological naturalism implies an underlying philosophy of philosophical naturalism (i.e., “there are no gods”):

Although Forrest wrongly implies that science can’t examine the supernatural, her overall argument makes sense. If you spend your life looking in vain for the Loch Ness Monster, stalking the lake with a camera, sounding it with sonar, and sending submersibles into its depths, and yet still find nothing, what is the more sensible view: to conclude provisionally that the monster simply isn’t there, or to throw up your hands and say, “It might be there; I’m not sure”? Most people would give the first response—unless they’re talking about God.

Gleiser’s agnosticism is of the second variety.

It’s quite curious, but understandable, that while Gleiser says that “I have no evidence for God or any kind of god,” he still says he’s an agnostic, and at the same time goes after atheists while keeping his mitts off of religionists. After all, it is the religionists who believe in stuff without any evidence, and Gleiser implicitly admits that. So why does he call out atheists but not believers? The answer is one word: Templeton.

Finally, what are the limits of science? Clearly they involve things that science can’t address because they involve issues not resolvable by empirical examination. Those issues are fewer than we think (for instance, science might be able to provide answer to “why do we find some things beautiful and other things repulsive?”), but we already know that science can’t fully resolve issues of subjective preference—like moral questions. If you decide what your moral preferences are (e.g., “it is best to maximize well being”), then you can approach the question empirically: what actions do maximize well being? But if that’s not your goal, and you have some other subjective morality (e.g., “abortion is immoral because it involves murder”), then science can’t answer that. (I do point out, though, that if the anti-abortion argument rests on the existence of a unique soul in humans, science might be able to tackle that.

I know very few scientists who would say that science can answer every question. It can’t, and sometimes we need clear-thinking input from philosophers and those trained to think logically. So when Gleiser emits the following bromides, I’ll just note that secular humanism, with no need for woo or “spirituality”, would arrive at the same conclusion:

. . . But let me play devil’s advocate for a moment, only because earlier you referred to the value of humility in science. Some would say now is not the time to be humble, given the rising tide of active, open hostility to science and objectivity around the globe. How would you respond to that?

This is of course something people have already told me: “Are you really sure you want to be saying these things?” And my answer is yes, absolutely. There is a difference between “science” and what we can call “scientism,” which is the notion that science can solve all problems. To a large extent, it is not science but rather how humanity has used science that has put us in our present difficulties. Because most people, in general, have no awareness of what science can and cannot do. So they misuse it, and they do not think about science in a more pluralistic way. So, okay, you’re going to develop a self-driving car? Good! But how will that car handle hard choices, like whether to prioritize the lives of its occupants or the lives of pedestrian bystanders? Is it going to just be the technologist from Google who decides? Let us hope not! You have to talk to philosophers, you have to talk to ethicists. And to not understand that, to say that science has all the answers, to me is just nonsense. We cannot presume that we are going to solve all the problems of the world using a strict scientific approach. It will not be the case, and it hasn’t ever been the case, because the world is too complex, and science has methodological powers as well as methodological limitations.

How you prioritize things like the value of lives versus the value of cars can be largely a subjective weighing of issues, but even then science can help resolve these questions once you specify an desirable trade-off between dollars and lives or dollars and convenience.  So no, science can’t by itself solve “all the problems of the world”. But it’s a damn sight better than religion in solving the problems of the world! Again, it’s understandable that Gleiser directs his ire at science and not religion. Why not indict Catholicism for saying it not only didn’t help solve the AIDS crisis, but made it worse? As Rebecca Goldstein told me, “Moral philosophy is a throughly secular enterprise.”

Gleiser is a Templeton flack, and I have nothing but contempt for his disingenuous attacks on atheism and science. Well, if he’s not disingenuous, he’s a very sloppy thinker.

Finally, why did Scientific American publish this article? The magazine is supposed to be about science, not science and woo. And the article is misleading and irrelevant, appealing only to believers who want to think their faith is consistent with science.

Here’s a photo of Gleiser from Australia’s Eternity News, which also had the headline below his photo:

h/t: Dave

127 thoughts on “Templeton Prize winner spouts more nonsense in Scientific American

  1. Grrr!
    “our present difficulties” – what exactly are these ‘difficulties’ that humans using ‘science’ have got us into?

    Talk about woolly thinking!

    I consider that the idea of a ‘god’ as religions conceive it, is absurd.

  2. And we can’t prove giant purple unicorns that shoot lasers out of their eyes don’t exist either, but I don’t see him standing up for them!

  3. When I see people talk like this I always wonder if they really believe what they are saying, or perhaps more accurately, has the money and notoriety they have acquired convinced them they are correct?

    As far a science goes, I say it is only a tool that tries to determine what is false and what is likely true. Unfortunately many important questions don’t break down as true/false, hence science can inform but cannot answer those. But science’s inability to answer those questions does NOT mean it is not a useful tool.

    1. I was thinking the same thing & concluded that he must’ve convinced himself of this stuff. Too bad.

  4. “We can’t prove God doesn’t exist, says award-winning scientist.”

    In the only sense of “prove” that his claim is accurate it is also a non sequitur. Science isn’t in the business of proving things, in that sense. In the more colloquial, common usage sense of prove that he is perhaps disingenuously counting on people to interpret his usage here as, science can, has, proved that there is no God. Provisionally, just as all of the results of science are supposed to be held.

  5. Gleiser’s remarks are frustrating and nothing more than what apologists have been espousing for years. What’s most troubling is that articles like this one encourage theists to continue using the tired “But there are scientists who believe in God” non sequitur.

  6. Nice clear take, PCC(E). I like the quotes.

    The headline writing is clever : “…,prizewinning physicist”. Mmmm. “Prizewinning physicist”. It’s true! Gleiser is a prizewinning physicist.

    Again, I think Gleiser assumes atheists make as big a deal of atheism as victims of faith make about religion. I personally don’t see atheism as something to wear in my sleeve, for various reasons. Thus I think Gleiser would be more productive if he could explain what is so important about faith.

    1. I just tried typing “prizewinning physicist” into AltaVista – ha! just kidding, Google, and the results are consistent with my view of the choice of headline, along with a separate expectation – the SA article is prominent in the “Top Stories”.

    2. I remember running across the article on reddit yesterday and I was surprised at a prizewinning physicist saying such a thing.

      Then I clicked on the link and saw it was the Templeton Prize.


  7. Delusional religionists are getting worse. Mike Pompeo just claimed that God might have sent Trump to save the Jews from Iran. How is anyone supposed to have any respect for U.S. international policy after a stupid remark like that?

    1. Yeah, they’re claiming Donald as the second coming of Queen Esther. And they’re playing right into the crazy right-wing evangelical eschatology.

      Meanwhile, these two corrupt momzers Trump and Bibi are humping each other’s leg.

    2. Sadly, I think people like Pompeo don’t understand that the rest of the world doesn’t buy that shit. Like when Trump went to the UN and everyone laughed at him.

  8. Gleiser’s remarks suggest to me he understands neither the meaning of atheism nor the meaning of the scientific method. A worthy recipient of a Templeton indeed.


    Many Templeton projects extol humility [see below Google page]. It’s a kind of ‘gaps’ argument – what we don’t know leaves room for the ‘spiritual & numinous:


    Marcelo Gleiser’s earliest interaction with Templeton that I can find, is this little ‘Initiative’ organised by Templeton in December 2010. Note the title! This is how Templeton trains their monkeys into using the language they want:


      1. Maybe he does the walk. The Dalai Lama does it best, and it takes a bit of practice, but I think anyone can do it if they really try.

        1. This makes me fantasize the Dali Llama doing the Woodcock courting boogie to Jimmy Cliff singing “You can get it if you really want”

          1. I have no idea how this got embedded, or the other videos, which I have no knowledge of.

            And the reply is to Yakaru.

    1. They probably get a little kit with their prize money that outlines how to talk to everyone on their junkets.

      1. Sounds right, but also he’s been vetted for ‘correct think’ by Templeton.org for nine years.

  10. “We cannot presume that we are going to solve all the problems of the world using a strict scientific approach.”

    Oh, Jeebus…THAT tired old straw man. I have never heard any scientist (or atheist) even suggest such a thing. We say only–and correctly–that nothing but science ever has or ever will solve anything. And certainly religion has nothing to contribute. I ask my “spiritual”friends this question:

    Can you name me ONE SINGLE THING about the universe that has been discovered by ANY priest/rabbi/imam/shaman/whatever, acting as such (in other words, through faith/prayer/revelation/anything but science), that was not known fifteen hundred years ago?

    1. Yeah, it’s like saying, “we aren’t going to repair this engine with this fork”. Therefore forks aren’t the answer to everything and whoever thinks so is a forkist and needs to learn humility.

    1. Oh that would be arrogant. He advocates humility which means he can’t proclaim such things dontchaknow.

  11. “Finally, why did Scientific American publish this article?”

    I did get a sense that the interviewer didn’t agree with Gleiser and, perhaps, thought hisw arguments were specious. That said, he didn’t go after Gleiser like I would have liked. Perhaps Gleiser would have walked.

    As to why Scientific American covered this, I’m guessing it’s because the prize was large and awarded to a “scientist”.

    1. Scientific american has become less scientific over time. When i was in elementary and high school in the 1950s and 60s (my father subscribed),the articles were written by scientists who wrote well. I could Always get something of value even if i only understood the first quarter or half of an article. More recently the magazine punlishes more articles by science writers who do not seem to tell me more than i already knew. While there are some good articles written by scientists and engineers, The overall committment by the editors to hard science seems to have waned.

      1. Yes, I couldn’t have put it better myself. I used to subscribe in the 70s and 80s. There were nice articles written by real scientists, often those that had done the fundamental work being summarized. Now, it is a shadow of its former self. Occasionally I will read something they place outside their paywall but it is never very interesting or inspiring. Of course things change. There are lots of interesting writing that’s free but you have to work harder to find it.

        1. Same here. I dropped Sci Am a long time ago after subscribing for 25 years or so. It just became shallow and vapid.

          I suspect Gleiser will be smart enough to hang around religious “news” outlets and other wooly thinkers. Hey, it worked for Deepity!

      2. Nice to hear the appreciation of Scientifc American from the Piel & Flanagan days. My father also got the magazine from the late 50’s through the 70’s and fortunately kept every issue. Even the artwork was wonderful. And Martin Gardner!

      3. Ugh. That’s too bad. I sometimes buy their publications on various things – they had a really migraine one a few years ago. I was subscribed for their online version but never had the time to read it. I too read it in the 80s as a pre-teen and teen and loved it.

  12. If Gleiser cannot even define atheist properly what kind of scientist is he? It is not that hard and it seems to confuse him like it does most religious people. Oh, maybe that’s it.

  13. “…to be respectful of people’s beliefs and reasoning, which might be community-based…”

    O Lord protect me from community-based reasoning.

    1. Dontcha know these days, everything’s about “creating community”; a phrase that’ll make me “throw up in my mouth” if I hear it again.

  14. To be fair to Gleiser, he probably didn’t attack the arrogance of those who are certain in their faith because he didn’t think they’d be reading Scientific American.

    I’m willing to bet that he’s part of the big fat humble little chorus singing “Oh, the atheists and the fundamentalists are both the same! They think they know without a doubt! I’m here in the middle, middle, middle which is temperate and true!”

    Well, everyone here seems to be in the middle, middle, middle and we don’t think you’re advising humility. We think you’re ignoring the case for the theory of Naturalism.

    1. Even Richard Dawkins admits that he can’t be sure god doesn’t exist. I’m of the same mind: I can’t say with 100% certainty that god or any other supernatural force doesn’t exist. I will never do so. I will only claim that there is no reason to believe in such things and, until presented with evidence (rather than the argument, “well, you can’t prove it’s not true,” I will continue to go on not believing the greatest extent possible short of complete certainty.

      1. His holier than thou (heh) “I’m an agnostic” was really irritating. By definition so are most of the atheists he is bashing as not understanding the scientific method. What a wanker.

        1. The only time I’ve ever encountered what someone would call “fundamentalist” atheists is in comments sections and message boards online, and those people are probably edgy teenagers and college kids who think they know everything about the world. I thought that when I was a teenager. Oh, how I wish I listened to so much of the advice from my parents and other adults back then. My life would be immeasurably better.

  15. You know why I never post on these stories? Because I just can’t bother. It’s too stupid. These arguments are just too stupid and so very, very tired. They’ve been made and rebutted over and over again. It’s not like free speech or something, where people can have legitimate arguments over policy.

    1. I wish you had been interviewing him for this article so you could have opened with that. 😀

  16. I cannot think of a single global problem caused by atheism, while among our most serious problems are those that stem from religion.
    And he is simply wrong that self-driving car technology will lead to engineering cars to choose who to crash into. The car will not choose any such thing once it recognizes the matter. It will simply hit the brakes. He is dreadfully/willfully ignorant of this well known issue.

    1. I disagree. Autonomous car systems will have to include logic that will, in effect, make decisions involving the value of a human life. For example, it is not hard to construct a situation where a car must avoid a pedestrian but, in doing so, will risk a crash that endangers the occupants of the car and/or other cars. Such a decision could be based on likelihoods of the pedestrian being able to get out of the way, the estimated risk of crashing and its danger to the occupants. Perhaps such calculations should value the occupant’s and pedestrian’s lives equally. However, it is not that simple. How does the death of a pedestrian weigh against the severe injury of five occupants? Does it matter the ages of the people involved? Should we value babies greater than seniors? Should we involve insurance and police issues in the calculation? Considerations might include the difference between doing nothing and taking action even when both result in equal loss. All of these things will need to be worked out.

        1. Yes. Of course insurance companies have had to deal with these issues for a long time. They are smart to not air their solutions in public. Most people don’t like the idea of putting a dollar value on such things as a human life.

          Long ago I remember seeing an article that compared multiple values for the worth of a human life based on various ways of calculating it. A low one is the value of the raw chemicals that make up a human body. Another is their value to the economy. Quite fascinating.

      1. I do not agree. Autonomous vehicles do not have to be able to solve the TROLLEY PROBLEM & it could be bad if we expected them to! An automatous vehicle manufacturer has a target of making her wheels some x factor amount safer than idiot humans – safe enough to save lives compared with the current system, safe enough to attract buyers, safe enough to minimise the cost of lawsuits & safe enough to get law makers/changers on board.

        Also the manufacturers of these intelligent vehicles are opening themselves up to all sorts of EXTRA legal problems if their products are designed to change direction and/or velocity based on a more-complex-than-absolutely-necessary interpretation of inputs. i.e. very deep cause & effect chains & value weightings of all objects in view. Here’s an engineer talking about it:

        A recent paper in the journal Science suggested that even regulation may not help: polling showed that regulation mandating such self-sacrifice wouldn’t be supported by a majority of people, and that they’d avoid buying self-driving cars as a result. That, of course, would result in far more deaths in the long run, as the endless deaths at the hands of incapable human drivers would continue.

        But to engineers at X, the Google sibling which is leading the charge to develop fully self-driving cars, the questions aren’t as interesting as they sound. “We love talking about the trolley problem”, joked Andrew Chatham, a principal engineer on the project.

        “The main thing to keep in mind is that we have yet to encounter one of these problems,” he said. “In all of our journeys, we have never been in a situation where you have to pick between the baby stroller or the grandmother. Even if we did see a scenario like that, usually that would mean you made a mistake a couple of seconds earlier. And so as a moral software engineer coming into work in the office, if I want to save lives, my goal is to prevent us from getting in that situation, because that implies that we screwed up.

        “It takes some of the intellectual intrigue out of the problem, but the answer is almost always ‘slam on the brakes’,” he added. “You’re much more confident about things directly in front of you, just because of how the system works, but also your control is much more precise by slamming on the brakes than trying to swerve into anything. So it would need to be a pretty extreme situation before that becomes anything other than the correct answer.”

        Even if a self-driving car did come up against a never-before-seen situation where it did have to pick between two accidental death scenarios, and even if the brakes failed, and even if it could think fast enough for the moral option to be a factor (Nathaniel Fairfield, another engineer on the project, jokes that the real question is “what would you …oh, it’s too late”), there remains no real agreement over what it should do even in idealised circumstances. A public tool released alongside the Science paper allows individuals to create their own ethical dilemmas, and the only consistent finding is that people are inconsistent – even when it comes to their own views. So it’s probably for the best that we aren’t trying to code those views into our cars just yet.


        1. I suppose I’m loosely letting “these systems” stand for “those designing these systems”, and it seems like they are “confronting” the trolley problem. (I don’t expect anyone ultimately to solve the trolley problem — the state of “no real agreement” has been with the problem since it’s start.) However, although the goal to “prevent … getting in that situation” is clearly what one would like to achieve, it in effect dodges the trolley issue. On the other hand, the decision “not to code these views” — which may well be the wisest course — is a response to the concerns trolley-like problems raise for automated systems.

          1. freiner:

            …although the goal to “prevent … getting in that situation” is clearly what one would like to achieve, it in effect dodges the trolley issue

            It ‘dodges’ [not an appropriate word IMO] the “trolley issue” because:
            [1] There isn’t an issue in the real world – it’s a philosophical construct that is not useful in autonomous vehicles
            [2] But, IF it was an issue it is [as you say yourself]insoluble [in all but the simplest cases]

            On the other hand, the decision “not to code these views” — which may well be the wisest course — is a response to the concerns trolley-like problems raise for automated systems.

            [1] It is the wisest course – we don’t want machines as proxy moral agents, because for the foreseeable future no program can be written [or taught] enough about the World & its values to operate except within very tight constraints: The road/lanes & objects that interface with the road/lanes. Suppose an autonomous bus is hijacked [or hacked] we don’t want machine moral proxies involved! All we need is that it’s well advertised that any vehicle which performs outside its expected envelope of speeds, routes etc is ‘blocked’ like a suspicious credit card transaction [in the bus case that might mean it pulls over & stops – can’t be reactivated until reset by a human authority]

            People & companies & politicians do not want machine moral proxies. For good reasons.

        2. Sure, that’s what engineers and project managers would say. I’m not saying that they directly code solutions to Trolley Problems. However, accident reconstruction and analysis will evaluate why an accident occured and sometimes find that the software made a choice in a way that can be interpreted by humans as having similarity to a Trolley Problem.

          The “slam on the brakes” answer is disingenuous and simplistic. Say version 1.0 of the autonomous car software implements a general “slam on the breaks” rule. Such a rule basically says that in no-win situations, go to the SlamOnTheBreaks() procedure. This will undoubtedly result in accidents and perhaps the car software will not be held responsible as it did the best it could. But then someone notices that it could have done something different that would have resulted in a better outcome. There will be great pressure to incorporate this fix into the code. Version 2.0 will do a little better than 1.0 but, at the same time, it will be a little more responsible for the outcome. This process will repeat while the computers get faster, the sensors improve, and the programmers’ skill gets better.

          I suppose at some point the public will say that cars are good enough or that some loss of life is acceptable. The cost of the accident analysis also plays a role. It will be interesting to see how this plays out.

          Something similar will happen with automated aircraft but, right now, we are still depending on the pilots. The situation with automated airliners differs from that of cars as the cost per accident is much higher. This will allow more to be spent on the automation as well as the accident analysis. On the other hand, trade-offs like the occupant/pedestrian one will occur less frequently with airliners.

          1. It’s got to be better than the current ape driving solution that results just over million deaths per year and terrible injuries.

            1. Indeed Diana. I am appalled (and often scared) nearly every time I drive by the number of drivers who do really stupid things. The number one, I think, is following the car ahead too closely. Many of the drivers I see would hit whatever is in front of them if something unexpected happened. That is the number one advice instructors at the BMW Performance Driving School tell you. I’ve been there for three courses. Among the other things you learn there that I’m sure most drivers don’t know are how to adjust your mirrors to minimize blind spots, look as far down the road as possible, pay attention to what’s around (and especially behind you), and how to use the anti-lock brakes. They do stop you quite well (and you can even steer the car with them on), but you have to be positioned so you can push the pedal all the way down, otherwise the ALB system doesn’t engage.

              I suspect a well designed AI system would drive at a reasonable speed for the road conditions, keep sufficient space around it that is could stop and could activate the ALBs if it had to, and wouldn’t be talking on an f**ing cell phone. It probably wouldn’t eliminate accidents, but it would probably cut way down on them.

              As an aside, in one course one of the students said “I watch racing a lot and you guys are always close to the car ahead”. The instructor said “That’s because there’s only one safe way to go thru a corner, and everyone knows what the track is like and what they’re doing. On the Turnpike, you can’t assume anyone knows”.

              1. As part of driving instruction, pupils should be shown what happens to them when the instructor slams on the anchors at 4 mph.

              2. I did that once to someone who refused to put his seat belt on in my car. He was siting in the back seat and before I did it I said, “Put on your seat-belt. If someone runs into us, I don’t want your adult body slamming into the back of me”. What was the best was that after he picked himself off the floor and put on his seatbelt, you could still hear me laughing maniacally. I even noticed my own laugh.

              3. Can you sing “Daisy Bell [on a bicycle built for two]” in a sad & breaky voice while plotting? 🙂

              4. All of what you said & people who don’t slow up when they see something happening ahead of them. A car loses control & so many people just drive right into the accident. I even remember as a young person, sliding in snow & people passing me. Back away, I haven’t gained control & can hit you!!

                I watched some of those mirror adjustment videos & have adjusted my mirrors on my cars – makes a big difference though I still have blind spots but it helps a lot.

                Many cars now have adaptive cruise control that adjusts your speed to be back from the car in front & it brakes when needed. It’s actually quite good. I’ve found my car’s cross traffic back up wonderful as it beeps far before I can see a car approaching me if I’m in reverse. I hate it when you’re in a parking lot & you can’t see to the sides because the vehicles around you are so tall. I always back up slowly for that reason, but some people insist and speeding through a parking lot.

          2. That’s right Paul – find a “slam on the brakes” quote when you know very well that’s not what actually happens in the world of Google X outside of a short interview. – then label it “disingenuous & simplistic”. Your style of argumentation is disappointing.

            The key point you swerved around so neatly [like a well behaved autonomous vehicle] is the guys in the interview said they haven’t the need to implement any kind of “trolley problem” solver into their code & they’re getting good results without.

            I say again: An automated trolley problem solver is NOT required & is in fact undesirable. Nobody wants it once they realise the consequences are thought through of such a concept. It’s silly.

          3. No. You are still holding to the view that the software made a choice in this case; that it recognized there was a serious consequence to either choice and then made a choice.
            The designs that are expected to come out are ones that I and Michael are saying (and Michael is saying much better:) If a car ever swerves to avoid a fat man to run over a lady in a wheel chair it will really be b/c it did not recognize the lady in time and could not brake in time.

            1. You misunderstand me. As I said, I would not expect the code to directly reflect a Trolley Problem calculation but the car’s behavior can still reflect logic with Trolley implications.

              I think your example is too simplistic. As you suggest, if the car didn’t recognize the lady then its decision obviously doesn’t involve her. The interesting decisions are the ones where both the lady and the man are recognized. Society may be ok with the car logic essentially throwing up its hands by braking as hard as possible and allowing the situation to unfold however it unfolds. On the other hand, if humans can see a strategy that improves outcomes, it will get implemented and may well involve Trolley trade-offs.

              One way the scenario might get complicated is if they train neural nets to make decisions on a case by case basis. They feed the neural net with a zillion scenarios, each labeled with the desired behavior. The resulting trained neural net will be presented with an actual situation and decide what to do. The labeling of such situations is done by humans and some of them may require Trolley-style reasoning to resolve.

              1. Paul

                [1] Please give an example of an autonomous vehicle trolley-style problem that needs solving. Construct a reasonable scenario.

                [2] Suppose a vehicle AI running on a neural net architecture makes the wrong decision & road users die that didn’t have to. How does one query the neural net AI about how it reached that fateful decision? Do you think neural nets can be queried?

              2. I’m not going to answer the first issue as I think I’ve adequately covered it. Constructing Trolley scenarios will get us into a battle over whether current technology handles the scenario or not. I have no idea which scenarios are currently handled by each autonomous car maker’s software.

                As to the second:

                1) When there’s an accident the neural net doesn’t handle as well as its human owners would like, the first question to ask is whether it was trained to handle the situation. If it wasn’t, then cases are added to the training data and the net is retrained or receives additional training. Note that the humans label the training cases with the desired outcome. This is where humans may have to consider Trolley-like problems.

                2) here are ways to query neural nets in the sense that one can analyze them to see why they produced the result they did for a particular input vector. This is actually a hot area of research in NN and Deep Learning these days. Since neural nets trained on millions of data points still make surprising errors, these techniques of visualizing what’s going on inside a neural net have been developed to help. Here’s a somewhat random link to the kind of work I’m talking about:


        3. I agree. And even if the software is written to calculate the odds in a Trolley Problem type scenario the cars systems are simply following the software instructions devised and vetted by humans. GIGO. Perhaps in some future time computers will be fully sentient entities that decide for themselves how to go about solving such dilemmas, but until then the dilemmas will be solved by humans prior to the computer having to possibly implement the solutions in a real life situation.

      2. The ‘car’ will not make the choice, computers cannot make such choices, they can only follow algorithms written by programmers.

        The algorithms or ‘choice’ will be made by lawyers, cost benefit risk analysts, possibly legislation and case law as it is made.

    2. A friend and I decided that self driving cars should have the ability to add a list of people you don’t like with a description/image/way for the car to accurately identify them. With this list you could dial up how much the car should avoid killing them. It could go from actively patrolling neighbourhoods to find the person and then running them down (full Christine) or just ignoring them and using some trolly problem logic as usual (full Herbie the love bug).

      This is why I’m not allowed to write requirements for things anymore.

            1. Astride a Harley? What would a rational thinker be doing riding a poorly designed agricultural pump from 116 years ago, and untouched by progress since?

              96% of Harleys are still on the road…

              …the other 4% made it home without breaking down.

      1. It’s quite possible that hackers will be able to do this in automated cars. Just take facial recognition software (or gait recognition, or hack the target’s phone and use its GPS to locate it, or …), give the location info to the car, and bypass its safety systems. Of course that last part is the hardest.

  17. Why do these clowns say science “has limits”? It would be fairer to say that science will come up against a wall — nobody knows what caused the Big Bang; nobody knows how life emerged — and that we simply haven’t figured out certain things. But this “has limits” trope is too close to saying that science is fundamentally ill-equipped to give us certain answers to certain things. Nope, not buying it.

    1. I should amend what I wrote by noting what Jerry said about subjective morality. But there’s no reason to believe that any religion or any “holy” book can successfully deal with moral questions or concerns [a nod to Goldstein’s great remark].

    2. I’d be more likely to accept “the universe has limits” in the sense that maybe we just can’t observe everything because we are just in the wrong space and time to do so.

  18. I would argue that any “test” for the existence of god would have the atheistic position as the null hypothesis. In this regard, atheism is not only compatible with the scientific method, it is fundamental to it.

  19. It’s always one god, isn’t it? We can’t prove that twelve gods do or don’t exist, but they never point that out.

    It is interesting that proof one way or another is implied to be a good thing, if you could get it. But that would make it science, and not having it makes it not science.

    No one knows anything about gods, and it looks like no one ever will. That is what atheism is about.

  20. “why did Scientific American publish this article?” Money.
    “why does he call out atheists but not believers? The answer is one word:” Money

  21. “The absence of evidence is not evidence of absence”

    This statement itself is anti-science.

    If P(H | E) > P(H), then it necessarily follows that P(H | ~E) < P(H). Absence of evidence is evidence of absence.

    If absence of evidence is not evidence of absence, then extraordinary claims do not require extraordinary evidence, Occam’s Razor is bunk, an unfalsifiable beliefs are genuine knowledge.

    All of those things are related. You can’t dismiss one without dismissing the rest.

    1. If evidence is absent because no one has looked for the evidence, or can look, then the statement is true. But if people have exhaustively looked for the evidence and there’s none there, then the statement is false. The absence of evidence is indeed evidence of absence.

      1. Right. If someone claims there’s a golden unicorn dancing in the air between us, and I can find no evidence of it, I’m perfectly justified to say “I don’t believe you.” Same with gods — all of them, as far as I can tell, although I’m rather fond of Ganesha, because he’s so cute.

  22. This reminds me of the most recent episode of Sean Carroll’s Mindscape podcast, an interview with Alan Lightman, a physicist/artist, but I think I’m overreacting to it.

    Lightman doesn’t like criticism of religious people; he thinks they deserve respect, that scientists have their own faith, and that some human experiences are so awesome they can never be understood by science (hinting at a spiritual realm, I guess).

    The thing that bugs me the most is the attitude that atheists should shut up and leave religion alone, as if religion is a benign, personal thing. (It’s a malignant thing that gets up into everyone’s face.)

  23. I must have missed this bit:

    I honestly think faith is inconsistent with the scientific method. What I mean by that is, what is faith? It’s a statement, a categorical statement that expresses belief in belief. “I believe even though I have no evidence for or against, simply I have faith.” Period. It’s a declaration.

    Of course such a balanced statement might have resulted in $1.4 million less funding.

  24. I find people often conflate “science” with “scientists”. The scientific method and the structures (more or less) that support it in academia work well for finding truths about the natural world. Scientists are just like the rest of us and since they are often in a highly competitive field, can be secretive about their work and even somewhat paranoid. They can also be nice people or real jerk just like anyone else. The can be arrogant or humble, etc. But these human foibles are irrespective of science. I think science is very humble – how could we have been so successful if it weren’t? Some fame seeker would simply proclaim something until it failed. We’d get nowhere.

  25. ‘If you spend your life looking in vain for the Loch Ness Monster, stalking the lake with a camera, sounding it with sonar, and sending submersibles into its depths, and yet still find nothing, what is the more sensible view: to conclude provisionally that the monster simply isn’t there, or to throw up your hands and say, “It might be there; I’m not sure”? ‘

    I’m pleased you said ‘provisionally’. It is a very big, deep loch and I suspect the task of searching every inch of it seems easier from an armchair than it does in real life. Especially when you factor in that the putative monster might not want to be found – it will see your submersible coming or hear your sonar and just go swim somewhere else. Just like most wild animals do.

    Also, of course, the ‘monster’ could be found tomorrow, mugging for the TV cameras, and we wouldn’t have to tear up one scientific principle; unlike if G*d were found to exist.

    So I think it’s a bad analogy. Would one equally conclude that MH370 probably doesn’t exist? (The search area was vastly greater but then, so was the effort put into the search).

    I will concede that I flip-flop on this. If the loch has been searched as thoroughly as the nay-sayers imply then the monster is improbable. But – has it? I genuinely doubt that.


    1. P.S. I deplore those Xtian fundamentalists who try to co-opt Nessie as ‘evidence’ of some biblical idea of theirs.

      As far as I’m concerned, the chances of Nessie existing is a matter of assigning probabilities, although that is a field full of uncertainties. Religion has no business in the matter.


    2. I’m one of those who hopes that something biologically interesting will be found in Loch Ness. Indeed I plan to spend a week this summer lurking there with a telephoto lens (and taking in the scenery)

      It is as you say a big lake. It is known that seals visit occasionally, coming up river from the Moray Firth, but there are very few good photos or identified sonar traces of them. Underwater visibility is so poor that a submersible would have to be very lucky to see anything.

      This is a good blog that keeps up to date with the evidence. http://lochnessmystery.blogspot.com/

      1. I’d love the monster to exist (and so, I think, would any biologist). Of course “nice” doesn’t equal “is”.

        I’d add a rider to the quote from Stenger –
        “The absence of evidence is evidence of absence—if that evidence should be there.” – AND if one should be reasonably certain of detecting it.

        Of the various searches listed in Wikipedia, I think only the 2003 BBC-sponsored one really qualifies. They had 600 sonar beams so, subject to the coverage being reasonably continuous, I guess that mostly satisfies my caveat.

        Also, last year (2018) there was a DNA survey of the loch which apparently should detect any ‘unusual’ species (results not yet out). I’ll leave it to the experts on this site to analyse that in due course.

        Have a happy holiday!


  26. “‘We can’t prove God doesn’t exist’ says award-winning scientist. Ohh that’s impressive! What award was it?”
    I imagine the headline belongs to the Eternity News, since almost any scientist will have heard of Russell’s teapot.

  27. As I said yesterday, Marcelo Gleiser and Adam Frank made the so called “science” of NPR unreadable for others. But I never read those two so deeply that I grokked Gleiser’s basis in mysticism. Of course we don’t know or expect to know everything, but meanwhile those vaunted limits of science are one of those things! Mysterious indeed – a decade ago we could not describe the universe but now we have a total description of mass, energy, and global process, at < 1/1000 uncertainty. Gleiser in hollow voice: "-m-y-s-t-e-r-i-o-u-s-".

    Anyone notice the new goalpost move on "scientism"? Seems Gleiser is not all to sure about the limits of science, so he now wants to make it about its use in society. But would I use science to chose the food I find appetizing? How would that work, Gleiser?

    why did Scientific American publish this article

    It is soft on religion, going the National Geographic way.

  28. “I know very few scientists who would say that science can answer every question.”

    I know no one.

    “here is a difference between “science” and what we can call “scientism,” which is the notion that science can solve all problems.”

    I’m scientistic.

    I believe firmly that there are no deeper answers than those discovered by (hard) science and science is the only way that humans can have (non trivial) justified beliefs.

    Sometimes science will lead to justified but false beliefs, to err is human, but other methods for producing beliefs are completely inadequate because their claims can only be true by accident. To use unreliable claims for justification is an error.

    This nicely solves my problem for justifying my atheism, but, as a consequence, I do believe other justifications for atheism cannot be justified.

    In my opinion Marcelo Gleiser’s mistake is to place a reliable belief forming process (hard science) on the same footing as a non-reliable one (human intuition) trying to justify his agnosticism.

  29. Gleiser’s writing poses science as if it is in competition with religion. This also seems to be JTF’s angle – the desired outcome is for religion to “enrich” science. However, that is a false dilemma. There is no competition. Religion is bogus. For science to have a meaningful dialogue with religion will be as productive as having a meaningful dialogue with alchemy, astrology, etc. That is to say, it will be productive and meaningful in ways that are unflattering to religion, at best, to say nothing about faith.

Leave a Reply