Feynman on group decision-making at Los Alamos (1985)
cs.au.dk
cs.au.dk
I think it's important for engineering decision-makers to be comfortable with this culture. It's disheartening when you have a great conversation like the one Feynman describes, the engineers are in agreement, and then the person in charge stands up and says, "Look, if everybody hates this idea, we can't be lazy and go forward with it just because we don't have a better one," or, "I guess we can't make a decision on this, because there's too much negativity and people are just going to nit-pick every idea to death!" No, no, talking through every possible objection to an idea is how you can tell that people are starting to zero in on it as the most likely best solution. The closer an idea gets to acceptance, the more thoroughly you vet it. The more it becomes clear that an idea is the best way forward, the more you start to prepare for dealing with the reality of it.
This makes practical sense, but not political sense, so it seems weird and unnatural to a lot of outside observers, who expect respect to be communicated by agreement, and emerging consensus around an idea to be expressed by a shared avoidance of mentioning the downsides.
Sounds a lot like the Braintrust described in the book "Creativity, Inc." which was accredited with a large portion of the credit for Pixar's success.
> This makes practical sense, but not political sense
The book says that to create the kind of environment that a Braintrust can function in takes a deal of care and one thing was all politics needed to be removed. Great book though highly recommend it.
One time I saw a room full of zoned out world-class scientists do this exact thing: reject the heavily vetted good option and accept the bad alternative because in the back of their heads they were just counting up positive attention and negative attention. The leader of the group realized what was going on, swung the conversation around again, re-framed things on more even footing, and people came to their senses.
Still, it was a pretty stark reminder that we're all human and prone to misusing the "attention survey" shortcut if the group is even slightly disengaged.
So there is something still missing from the recipe for coherent decision-making. It's the reason Feynman says these are "great men" - it's NOT the process which makes it work, but rather a quality inherent in the character of these men.
I don't have an answer for what that may have been, but in the spirit of your second paragraph I just wanted to bring up this thought.
I've characterized this before as the single most important aspect of science [1]. It is obviously very important in engineering. It's important in navigating the noise of propaganda and lies out there in the world. It's important in day to day life, e.g., it's hard to scam someone who is actively trying to figure out why your scam offer is not going to work rather than helpfully talking themselves into going along with it.
I suppose there is an excessive negativity trap. In the end you must decide something. But for many people that would still be an upgrade, vs. letting their natural inclination to self-justify every decision with the full power of the human brain run rampant.
So instead of looking for corroboration, they actively set up a situation which they think could disprove their hypothesis. This is a very economical way to zero in on truth.
It's also why I ask seemingly stupid questions when someone teaches me something. It's not that I don't think I know the answer -- it's precisely because I think the answer is obvious. One out of fifty times the answer is different, and that tells me my mental model is seriously flawed and I should question everything else I thought I just learned.
> Of the fifty-odd biases discovered by Kahneman, Tversky, and their successors, forty-nine are cute quirks, and one is destroying civilization. This last one is confirmation bias - our tendency to interpret evidence as confirming our pre-existing beliefs instead of changing our minds.
I certainly agree with the premise that there is a deep epistemic flaw in the nature of our reasoning that we can counteract with conscious intervention, such as your suggestion.
[1] https://astralcodexten.substack.com/p/book-review-the-scout-...
On the other hand, people who are super confident and wave away any limitations of their approach have invariably been problematic hires.
There is a cultural aspect to this too, because America for some reasons worships confidence and hates ambiguity.
I just started reading Boswell's biography, and the introduction to the Everyman's Library edition makes the same point:
> Part of this comes from a literal-mindedness that won't mistake one fault for another or all others. This worked even when his dominant feeling was not of fondness (as for Smart) but of dislike (as for Sterne). Boswell reports Goldsmith toadying up to Johnson, in a conversation on Sterne, by saying 'And a very dull fellow': Johnson replied 'Why, no, sir.' Fact.
I feel these meetings work best when there is no career pressure, but when people have aligned goals and just want to make progress. It's easier to leave good points unsaid if everyone prefers having any conclusion to having no conclusion.
Second is the chairman facilitating who didn't controlled the conversation, didn't leave time for bullshit, and probably had a very tight agenda.
Maybe being alone at the top is good for the ego, and some people like that, but it is also boring. If you are at the top in some field, is is probably your passion, something you want to talk about all day, but no one understands you, you feel like you are talking to babies, repeating "obvious" things over and over. On the opposite, you can't find anyone to teach anything new to you, being a pioneer is nice, but if you get stuck, you can't count on anyone but yourself, and it can become tedious.
So, when you finally meet someone on your level, it is someone you want to keep. At last, you can talk with someone who understand you, who can provide a meaningful input, and even teach you a thing or two. And even if there is some rivalry, you have a worthy opponent, not an annoyance.
Don't hesitate to call others out for their behaviour, as it occurs. And involve those who seem to be withdrawn or quiet, ask them about their opinion and input. It takes concerted efforts, if not coaching and training, to get out of that habit.
Perhaps sufficient greatness of attention span is a prerequisite to remember what everyone has said since the start of the meeting :-)
An interesting thing happens when there's a high level of respect.
People stop being so anxious about not "winning", or "being heard", people just remove themselves from their ideas and feel comfortable and confident.
I left that company because the company around that team was a bit shit, but I wish I could foster an environment of high trust and mutual respect like that again.
Whether people worry about being heard, whether being winning is important or not influences a lot how they will discuss. Whether you can afford to remove yourself from the idea is artefact of the environment too.
I think this is a critical component of any moonshot - you have to have competition or some motivation that feels like an external compulsion.
- nuclear threat is a serious cost to productivity and well-being - the bomb merely brought forward a conclusion that was already looking inevitable
There's an obvious counter that some nation likely would have developed nuclear weapons independently - America was merely the first.
Nonetheless, I think it's not a given that the bomb was a positive.
Another factor was that the U.S. thought it was very possible we’d be going to war with Russia and was slow to demobilize so there was definitely the intent to intimidate the Soviet Union even that early.
Sometimes there are only bad options available. The bomb was the best choice
"After the war, the commander of Japan's minesweeping operations noted that he thought this mining campaign could have directly led to the defeat of Japan on its own had it begun earlier. Similar conclusions were reached by American analysts who reported in July 1946 in the United States Strategic Bombing Survey that it would have been more efficient to combine the United States' effective anti-shipping submarine effort with land- and carrier-based air power to strike harder against merchant shipping and begin a more extensive aerial mining campaign earlier in the war. This would have starved Japan, forcing an earlier end to the war." [0]&[2]
"While some of the main actors (the UK, the US, Germany, and Japan) escaped famine..." [1]
"In Britain, Germany, and Japan, rationing was effective; even though food availability was sharply reduced almost everywhere, few literally died of hunger. Britain and Germany managed to maintain food consumption per capita at about 3,000 calories throughout the war. In Japan consumption fell from a norm of 2,000 calories per capita before Pearl Harbor to 1,900 calories in 1944, plummeting to 1,680 calories by the war’s end." [1]
0: https://en.wikipedia.org/wiki/Operation_Starvation
If you mean developing the bomb, then fine.
If you mean dropping the bomb, then that seems pretty controversial.
I thought it was pretty well established that the USA knew Japan was willing to surrender, but dropped the bomb anyway?
This is the opposite of what I usually hear? I've heard things about how Japan was getting ready to fight to the last man, starting to arm civilians.
I have heard that the emperor was maybe willing to surrender, but that the country was essentially being run by the military leadership, who would ~never see surrender as an option.
If the USSR decided to invade Japan on its own, who cares? We could let them fight. Or we could ally with Japan.
I honestly see this as a purely gut-level reaction to the astonishing power of the nuclear bomb more than anything about the victims. The casualties of Hiroshima and Nagasaki are but a rounding error next to all those who died by fire bombing, carpet bombing, siege and starvations, and all the other infernos of WW2.
It's not that this makes it okay, it just makes it strange that the bombings get all this attention in popular culture over other much more horrific atrocities and massacres. This was probably an effect of the 1960s/1970s/1980s terror of nuclear projected back on the events of 1945.
The A bomb is extensively discussed late in the series, with shift in the strategy from strategic bombing to mass firebombing as US were preparing the invasion of mainland japan.
TL;DR: > I thought it was pretty well established that the USA knew Japan was willing to surrender, but dropped the bomb anyway?
Simply not true, however answering if Abomb was good move or not, is such a complex and nuanced question you cannot just have a simple statement for or against.
The USA feared the invasion of mainland japan because of the unimaginable looses they would incur. Japan's population was prepared to fight USA with sharpened bamboo stick, and I mean population as in civilians, as in EVERY single person is possibly out to get you. Just imagine what kind of mental damage you would put of your soldiers to have to mow down scores of suicidal civilians. Note that this behaviour was already showing up as US were nearing closer to actual japan.
Also you had USSR freed from fighting nazis having a massive score to settle with japan. Do you want uncle stalin also be involved? (given what happened to west/east germany was it better japan avoided est/west japan split?)
Have a listen to [1] on the accounts of firebombing of Tokyo. It horrendous fate in comparison to A-bomb. On top the horror of slowly burning alive trapped in fire cage there were far more death from firebombings/carpetbombings than Abombs.
[0]https://www.dancarlin.com/product/hardcore-history-62-supern...
[1]https://www.dancarlin.com/product/hardcore-history-67-supern...
There are a few different theories about this, and still some controversy about it all to this very day as I understand it.
One idea holds that Japan would have surrendered earlier had the Allies not insisted on the abolition of the role of the Emperor (who as I understand it, is something akin to God in Japanese culture). This was apparently a step too far for the Japanese at the time.
Another idea holds that even if the Allies had invaded, the Japanese would have quickly capitulated IF the Russians were part of the invasion. Supposedly Japan was uniquely afraid of Russia at that moment in history, for some reason. I don't recall now the details behind this suggestion.
I can't say much about the veracity of either position, but those are two themes I've heard repeatedly about the situation.
If you have any evidence otherwise, I would be interested in reviewing it.
By the time nukes were dropped on Japan [0] in August 1945 nazi Germany was defeated, the war in Europe ended, Soviets attacked Japanese army in Manchuria [1] on August 9-th -- the day of Nagasaki bombing.
Dropping bombs on Japan had one main purpose - to show the Soviets who is the boss and prevent them from invading and occupying Japan before the USA can do it.
[0] https://en.wikipedia.org/wiki/Atomic_bombings_of_Hiroshima_a...
Between that, and the amphibious attacks against Korea, the Kurils, and Sakhalin, the US was on a pretty short clock to end the war before communism had a chance to move in, and settle down.
Stalin respected the pre-agreed partition of Europe, but neither Postdam, nor Tehran said very much about who would be in control of what part of the Japanese empire after the war... And in practice, this meant that when the war ended, possession, in the form of 'whose army is parked where' was nine tenths of the law.
The United States was not at all interested in seeing an East/West Japan partition. North/South Korea was bad enough.
In so many normal conversations people are repeating over and over again because they assume it goes in one ear and out the other and they are right.
Another part of the issue is the internet, for many reasons, but specifically related to this is that it's killed people's curiosity and their ability to discuss a topic for fun. The internet gives people their opinions, and answers any ambiguity in an instant. Want to know what the best tire for your bike is? No need to discuss it, the internet told them what's correct, case closed.
It's rather curious to me to see how others live their lives in ways that they're limited on details that I take for granted as normal and regular. I see many people here describing activities that I haven't done in years or ever too, and then run into something like your comment above, where sitting down with a friend for a drink and a few hours of conversation is a rare thing, though in my case I usually find the time to do it at least once a week to refresh and relax, and have done it without interruptions for years now. Just an observation on these differences between lives in general.
In my workplace during the pandemic I have seen this bias towards action play out. New (barely MVP) systems and integrations build in a week that would have been a year to get to the same point normally. A yes, there were a few nonsense "$action must be done" as well, but we managed the crisis too.
There has certainly been more discussion & review of options than in the linked article, more debating of relative merits. But, then again, the circumstances I've been working through in my organization are not as clear cut as the optimal method of extracting a substance where there are discrete known facts. And the people involved have been experts in varying domains, not all physicists or related backgrounds where the merits of an opinion can be understood by the others without more discussion. Also this was one meeting: I doubt that every meeting among the people in the article all went so smoothly, with never debate or need for someone to rearticulate or expand their explanation of a point of view.
Please note that I am not saying these men weren't great, and that the group dynamic Feynman describes isn't at least partly indicative of their level of intelligence. What I am saying is that the group dynamic described is not unique to people of towering genius. Having weathered many crisis (to be sure none on the scale of WW II) my observation is that this group dynamic cannot be disentangled from the crisis context and only-- somewhat simplistically-- attributed to "these were great men". Greatness of mind is not a universal panacea for indecision and uncertainly. Crisis on the other hand forces people to push through those things.
I've seen it a lot in meetings, part of it is excitement, people really eager to bring their Knowledge to the table, or that they need to get things out of their system before the conversation goes somewhere else.
I think there should be a lot more formalized meetings, in the sense that people are coached to finish one sentence / idea before moving on to the next, to ensure that all voices are heard, etc.
I worked at a mediation center with a woman (a founder) who could listen to two people talk about a dispute for 90 minutes, not take any notes, listen carefully and make a detailed verbal summary of what the people said. It amazed me.
I was in a call just now where there was a knock-down drag-out discussion going on that we've had before. It was tempting to disengage, daydream. I was waiting to write this comment the whole time but was able to stop it because it was worth listening to the group which is the only way out of the impasse. I find myself asking, "What is it going take for this person to feel heard?" (I wish I could just ask them that!)
David Snyder is a hypnotist who makes videos on Youtube and he says that if you want people to feel mirrored it's powerful to remember and be able to repeat what the person said using the same words in the same order. Even if there is a place for paraphrasing, the more exact you can be the more certain you can be heard.
We were having a stoned discussion at my house and my son was feeling that people were talking over him so I cleared off the whiteboard and summarized what he said. I was not going to "lose" and let him walk away saying that I hadn't listened to him. Sometimes he got really upset when I made him pause so I could take a picture of the board and clear it but I didn't stop until he didn't have anything more to say and he acknowledged that he'd been heard.
I have an ideal of working towards of being able to offer people an unreasonable level of mirroring because of the power it has for drawing people in. I know I can't be perfect at it and can't expect perfect mirroring for others but I try to imagine there's no limit to how good I can do it.
I was at a party the other day and a woman tried to flatter me based on a superficial thing about my intelligence and my first reaction was "Yuck! I've always felt my intelligence was something that separated me from other people. I'd appreciate it if somebody said I was funny or brave and probably not be able to think straight if somebody said I was handsome, but..." I had to step back and try to appreciate the sentiment but I thought about it and thought some more and decided I'd rather not make that mistake when I'm trying to seduce someone.
I see a lot of people at the gym who have their bodies on display but are really defended and I don't feel I could talk to anybody about their superficial characteristics. I am only going to talk to somebody if I can show I really saw something about them that was personal, unique and touches me.
There's one lady who really whoops it up when the Baltimore Ravens are playing on Sunday and that's one of the very few people I can talk to because boy I know she is crazy about football!
And countries.
To top it off the chairman of the day summarised the whole event in a speech that I still remember to this day. It was so charming, intelligent and erudite. I was floored.
The parts on Brazil were particularly interesting.
If you've worked in a bureaucracy, you'll know that the important decisions are not made in meetings.
I think the notion of "great man" like the genius of "Einstein" or "Newton" is seriously skewed. If you take a closer look at their correspondence and upbringing, they both were very well "connected" in their own ways beforehand.
This picture of a loner sitting bored at a Patent Office and all by himself coming up with 4 groundbreaking papers (photoelectric effect, brownian motion, special relativity) in 1905 - [or Newtons "annus mirabilis" after fleeing Cambridge University because of the plague in 1665 and "quarantining" himself for 2 years in his native village (by his own account (!) there he singlehandendly developed integral calculus, dicovered the composite nature of light and finally refined his gravitational theory)] - readily highlights the individual instances of "genius" which I do not want to dismiss here - but entirely blanks out how those free floating ideas were able to circulate and bounce off each other giving rise to those great insights in the "quiet" moments of those "geniuses".
Only after their recognition of "genius" they seemed to develop some really unhealthy stubbornness.
Newton - beginning in 1686 with his publication of the Principia - "disputing" any influence Hooke might have had and at his height of power letting his "Royal Society" create a comittee to decide against Leibniz "priority" regarding "calculus" in 1712).
Einstein beginning with the Bohr debate (1927 onwards) and productively contributing to the EPR paradox in 1934 (further developed by Bell to the infamous "Bell's theorem") felt more and more "misunderstood" and isolated himself to work on his own "unified field theory" and from then on did not participate in any of the "quantum mechanics" of his time.
Yeah knowing he wrote it, I guess it does come off a little more egotistical than I at first assumed. Still a good book though.
Feynman was humble in that he tried to learn what was true instead of assuming that he already knew. He was always looking to improve his understanding.
Being humble in the search for truth can sometimes even seem arrogant socially. It means that you'll regularly question not only your own beliefs but those of other people as well.
Although according to Murray Gell-Mann, Feynman might have gone a bit too far in that regard
Like, being humble, competent in your field, mature enough to actively listen to other people, etc, all help. But there are also more subtle factors like compatibility, company culture, project parameters, etc, that are also necessary to have that dynamic where everybody cooperates in discussions.
Getting those subtle factors to align is hard, sometimes it's down to conditions you can't control, and it's definitely not as simple as "just" getting great people in a room.
https://www.youtube.com/watch?v=e3msMuwqp-o&list=PLctkxgWNSR...
In the end, you’re left committing to only one idea and everyone is supporting it because that’s the only way it’s going to be successful. This is known intrinsically by everyone participating.
What this quote is missing is all the context. They were fighting for time and had to make a decision to get started on experiments. Many of these contributors were doing experimental work and Feynman was doing theoretical work.
Also he was 24 years old at the time.
The best teachers never really quit being students.
Although an anecdote, I would imagine there are plenty of examples where the opposite would happen to people who could still fairly be called great men/people.
In fact, some times, great people are necessarily obnoxious - or at least cold. I can't imagine that kind/warm people could have got Tesla or Apple to where they are, it takes a certain hard nose to make hard choices in business.
Perhaps the surprising bit was that they could all agree on the not-ideal method because they needed to move on and the meeting was a formality of agreement and alignment, not a brainstorming fight for your own idea session?
We have another anecdote of Feynman being shown blueprints of a nuclear power plant and he doesn’t understand and eventually asks a stupid question about a part, “expecting” to be laughed at and told it’s a door, but it turns out to be a valve which could cause a problem if it fails. Nobody else had noticed, implying they were pretending their understanding, and they left thinking he was a great insightful man which he’s saying he isn’t (but is likely downplaying for comedy as well as humbleness)
When I imagine a manager, my imagined ideal is a warm and conscientious person.
For individual contributors, I imagine lukewarm and single-minded. For team members, warm and cooperative.
So if most of humanity values warmth, how did executive types end up at the top of the pile? I thought that I would eventually move into executive levels when I was focused on a management career, but the managers I admired who could move their people always seemed distinct from the hardasses who could move mountains. I'm just now realizing this.
Anyways, your comment piqued my interest because of the big impact the mental model of "warm/cold" has had on my social life. I'm more compassionate of cold people because I assume that they just haven't learned that mental model. I also literally sometimes imagine myself as cozy warm to help me open up and welcome in.
Another interpretation might be that he considers this a rare trait in people.
I was lucky enough to work with some 'really really clever people (math+physics+cs)' in my career. Bar one or two generally the nicer they were the higher their intelligence seems to be at least from my point of view.
The reverse were also true. I've also had the misfortune to work with some really 'difficult individuals' and you can spot their lack of basic knowledge + intelligence from a mile away.
Maybe it's an evolution thing ? The "dumber" you are... the more you need to be aggressive (I.e Caveman: No charm, but I got a club to seduce females + competitors) to have your genes continue.
Versus if you clever, you don't need all that aggression, you can finder clever ways to "woo" the women to carry on your genes :) ?
Most medical group meetings are:
a)A small number of people that understand the issue (generally specialists) being drowned out by:
b)A large number of people with a vague understanding of the issue (related specialists and generalists)
c)A number of people with no understanding of the issue with the most political clout (administrators)
And then a final decision made by the least knowledgeable, most political administrator with the most vested interests.
Can we attribute this to the fact that in physics, everything can be validated mathematically? Hence less room for debate.
There’s self section of individuals with high working memory into such roles. There are many managers who attend meetings all day and can’t synthesize what’s being discussed in real time, indicating that this isn’t about practice.
Sad if most companies are not like that, I must have been lucky to be in the right teams then.
But also it's important to separate oneself from ideas that appear in one's head and not be too closely invested in them.
If the committee didn't choose Compton's idea, it would just be an average lame committee where quality of thought isn't a priority.
There's a reason "design by committee" is an epithet. People thinking in groups normally suck at it.
Meaning no gathering of material for self-aggrandizing or as ammunition for department warfare.
No PR-stunts and no humiliation for gaps in knowledge. The person and ego has to be suspended for a moment and the "competition" of man go take a hike, replaced by a competition of ideas.
Fairness has to be upheld, as in everyone in the round has to be a good sport and accept his/her idea to be less likely. Todays self-obsessed culture makes such a round nearly impossible, except - ironically for near anonymous internet forums.
I really dislike the wording of that title.
Whatever politico has enough sway with the social media companies makes it known what the proper narrative is, and then any other point of view is simply labelled as misinformation.
That way we don't have to put up with the misinformed views of great men.