Strong Opinions Loosely Held Might Be the Worst Idea in Tech
blog.glowforge.com
blog.glowforge.com
Not "I shout the loudest about it".
And "weakly held" meant "I am willing to believe there's data out there showing I'm wrong, and I will look at any new data with an open mind". Not "I'll flip if somebody shouts louder".
And sure, on the face of it, the proposed solution here addresses that, but... we don't know "how sure" we are of something. It's a made up number. And I'm willing to take bets that the same people who do the loud shouting will be the ones who just declare every certainty 100%. (The smarter ones will pick 98%).
The key to making this work is openness to new data.
If you want a rhetorical trick to facilitate debate, I prefer the old "Tell me what I'm missing" - it presupposes you don't have all the data, nudging you to be open to counterpoints and validating people who might want to offer counterpoints.
Not as much as you'd like. For many, "your job to push back" is saying "Your job is to argue." Many people want to discuss, not argue. Regarding "I want your criticism":
I took a communications workshop, and we discussed this. Make a statement/claim, then follow it up with "Let me know if I got anything wrong." Half the class thought it was a good idea, and the other half said it wasn't (and likely wouldn't engage). The reason was that too many of them had seen the tactic used by insincere, argumentative people, and the very phrasing comes off as adverserial to many (like "show me how I'm wrong!").
The other half of the class was completely surprised by this ("When I say it, I'm sincere!").
The phrase that didn't trigger any problems is "This is how I see it. What do you think?" or "This is how I see it. What is your perspective on it?" or even better "This is how I see it. I'd like to hear your perspective."
"Strong opinions, weakly held" (in it's best spirit) is often a crutch for people who feel paralyzed by ambiguity. This mindset lets them sweep uncertainty under the rug and continue operating ("bias for action") till feedback requires them to change their mind. The problem with that assumption is that the willingness of people around them to express feedback is inversely proportional to the confidence they express.
On the topic of being comfortable with ambiguity, I feel Feynman puts it best: https://youtu.be/_MmpUWEW6Is
I've been heading this way over time, and as you clearly have experience that it works I'll make an effort to try harder. Nothing like learning from other people's wisdom - thanks!
But there's a valuable philosophy locked up in the meaning I've always attributed to this saying. Namely, that many--maybe most--architectural decisions in tech don't actually have a correct answer. And so it's more important to pick an approach and stick to it, unless and until there's some compelling reason to change strategy. BUT, when that reason does appear, follow it, and change your mind.
What he's talking about looks like jerks using a saying as an excuse, I've seen the same thing with "direct/telling it like it is" as a shield to be an asshole. That has no place anywhere.
It's toxic to be unequivocal in situations that are inherently ambiguous and require input.
Also, there's something powerful about orienting towards action; often, what matters is less architecture but simply moving forward at a reasonable pace. A mediocre decision is usually better than no decision.
The catch is that people will not agree with what is ambiguous and what isn't. Don't assume that something that seems unambiguous is.
The other thing not mentioned in the article: The likelihood of misunderstandings and miscommunications. The issue could be as black and white as "This functions gives the wrong answer with this input." It's simple, right? We both agree on the answer it should give. Just run the function and see. How much more unambiguous can you get?
What I've found is that a significant percentage of these discussions aren't about what I think they are. Yes, I can insist that the function gives the wrong value. And no, that's not what the other person is even discussing, and neither of us has realized that we're both talking about different things. Stating it unequivocally is a barrier to the discovery process for realizing we're not on the same page.
Some people are good at verbal debate. But many don't revel in it, and it's always a pain point for them. They don't want to work with a constant debater. I know in a lot of tech culture, we set the expectation on them to speak up, but that culture is working against human nature - and often results in losing talent that doesn't fit in.
Don't make assumptions about what we assume :).
There are innumerable situations wherein there is legitimate experience and authority for people to make fairly clear and unambiguous decisions.
It's patently wrong to suggest that unequivocal positions are inherently toxic.
This is why small multidisciplinary teams work well. Where multiple diverse opinions come from different domains but there's clear leader in each domain.
From there it can get reviewed by other teams or layers so ideas get challenged and prodded until it gets made better.
Rather having one big group of people talking and having the loudest people the most time.
Proper delegation to smaller teams, ownership, feedback loops, and being deferential to domain experts can all help resolve this problem. Which also means hiring good people you can trust.
I'm reading your comment and kinda wondering what it has to do with my comment that you're replying to. I never suggested such a thing.
>There are innumerable situations wherein there is legitimate experience and authority for people to make fairly clear and unambiguous decisions.
I'm not disputing that. The context of my comment is the discussion around it. My question is: If that is the case, does there need to be any discussion around it at all? Just make the decision. It kind of gets to the point of the article. If it's unambiguous and you're sure about the decision, you can present it that way and close any room for disagreement. If you actually want people to express disagreement, then it means you are willing to consider your analysis of ambiguity would be wrong.
My counter would be:
There are innumerable situations where people's belief that the situation allows for clear and unambiguous decisions is wrong.
I'm not saying your scenarios don't exist. I'm arguing about Type I vs Type II errors.
In terms of logical necessity, no.
In practical terms, that's less clear.
> In some situations, people have experience and know better.
The problem is that the people you deal with aren't robots; being unequivocal even then can both make it harder to get people to accept your conclusions and be toxic to relationships.
Projecting your own certainty that you know better than everyone else is often counterproductive in convincing people that you do, in fact, know better on the issue at hand.
Choice #1 is debating with engineers who have strong but loosely held opinions. This path is the way of 'moving fast and breaking things'.
Choice #2 is debating with engineers who have mild convictions at best. This is the path of endless bike-shedding and yak-shaving.
Option 1 implies that people aren't afraid to take leadership of an idea and accept responsibility for its consequences if things go wrong, which is conducive to iterative development and innovation. It implies that the person making the strong case for a decision has done thorough research on the matter and has good reason to be confident in their approach.
Option 2 is more of an ego-preservation mechanism where people hedge their opinions to protect themselves from criticism if things go wrong. You don't know how they really feel about the matter because everyone is walking on eggshells in order to preserve one another's egos. People don't want to say anything with conviction and most of the time would rather say nothing at all.
Anyone who has ever tried to elicit candid feedback from option 2's engineers will know what I mean.
>Option 1 implies that people aren't afraid to take leadership of an idea and accept responsibility for its consequences if things go wrong, which is conducive to iterative development and innovation. It implies that the person making the strong case for a decision has done thorough research on the matter and has good reason to be confident in their approach.
Yet I've come across too many people who have identical behavior to Choice #1, and have not done enough due diligence in their research, and whose reasons are strongly biased towards their own comfort.
>Option 2 is more of an ego-preservation mechanism where people hedge their opinions to protect themselves from criticism if things go wrong. You don't know how they really feel about the matter because everyone is walking on eggshells in order to preserve one another's egos. People don't want to say anything with conviction and most of the time would rather say nothing at all.
Yet I've come across too many people who have identical behavior to Choice #2 who don't have any egos at stake - they just do not speak in a confidence signaling manner. Their tone is no reflection of the accuracy of what they say.
>Anyone who has ever tried to elicit candid feedback from option 2's engineers will know what I mean.
I have, and have often succeeded.
Anyone who has tried and failed with option 2 is strongly encouraged to study the art of communication. I assure you: The failure is not on the part of the people who exhibit this behavior.
Your description of the situation is incomplete, and is missing quite a few other Choices. Amusingly enough, in the book Crucial Conversations, they start off early with people who have the same view as you on Option 2. They call it The Fool's Choice: "Either I state what I think and damage the relationship, or I preserve the relationship at the cost of not getting what I want." They discuss it early in the book because people will simply not become better communicators while they believe in the Fool's Choice.
After reading 2-3 different books on communications, you'll generally see communication problems in the wild as, frankly, textbook examples of bad communications. Almost all scenarios are in one of these books.
What can I say, I practice what I preach ;)
I'd like to point out that your comment has many characteristics of a "strong opinion" as well, which I appreciate. For example, strong statements "i.e - I assure you: The failure is not on the part of the people who exhibit this behavior" and research based evidence. It will be of use to me in formulating a judgement on my own strongly expressed opinion. If you had instead said something like "The failure is not on the part of the people who exhibit this behavior... but this is based solely on my own personal experience and might not apply here", it would have been a much less convincing argument leading to dead-locking opinions at best.
With respect to your point about The Fool's Choice, I would think that in an ideal world, stating what one thinks about a technical matter should not be damaging to a relationship at all (unless truly fragile egos are involved). One should be free to state strong opinions in the best interest of the organization, but not be afraid to back down in the face of contradicting evidence. It should also not be about 'getting what I want', but 'getting what's best for the organization'.
Well, I come from an academic background, which is rife with debate. The style you prefer is natural to me.
And yet, despite that, I had to think and see whether it is worth the effort to engage with you. If someone who enjoys debate has that thought, then people who do not enjoy debates are even less likely to engage with you. That's the point I'm trying to get across.
For me, debate has a low cognitive load. For most humans, that's likely not true. Yet for most people, simple discussions have low cognitive load. So in general you'll get more out of people if things are not framed as a debate. Putting the onus on the other party to counter is framing at as a debate and is working against human nature.
I also want to touch on:
>but this is based solely on my own personal experience and might not apply here", it would have been a much less convincing argument leading to dead-locking opinions at best.
It need not lead to deadlock. You're allowed to disagree with people. You're allowed to say "This sounds like your world view, which doesn't match mine. Do you have any resources to support your view?" (Also, bearing in mind whether you have any to support your own). And at the end, "It seems this is a matter of opinion, and at the moment I am going with X. Thanks for your input."
(Not saying it's easy to talk like that if not used to, even I don't do it all the time - or even 50% of the time - but people do learn it and become good at it).
And even worse than a deadlock is people simply electing not to give you their opinion, an option I almost took.
One last tangential point. People in the HN crowd tend to have very strong analytical skills. One thing I've learned to be mindful of is if someone doesn't have good arguments for their position it does not mean they are wrong. Lack of good arguments from a person is a very poor signal of whether they are correct! I learned the hard way when on a number of occasions, I turned out to be wrong, and yet someone had told me the correct thing - they just were not good at articulating their reasons. I've realized that, time permitting, it is my responsibility to explore their suggestions. I cannot hide behind "Yes, he didn't explain it well."
>With respect to your point about The Fool's Choice, I would think that in an ideal world, stating what one thinks about a technical matter should not be damaging to a relationship at all (unless truly fragile egos are involved).
That is what the Fool's Choice says: One should believe they can state their perspective without hurting relationships. The problem often arises in the purpose or intent. Is your goal to state your perspective, or to get feedback on it? What you've described is the former. Is your goal to tell people, or to have a discussion? For a lot of people, unless it is clearly phrased as the latter, they will assume the former, even if it wasn't your intent.
We don't live in an ideal world.
(Off topic, but I had a manager who often said "In an ideal world, we would...". At some point I cut him off and said "In an ideal world, I wouldn't need to work". Ideality has its degrees.)
It is generally overread and therefore fairly neutral in terms of click-through. In cases where integrity is given up on, you won't find weasel words, e.g:
I'm fairly sure that you meant to say He should change it to: "I am 75% sure that 'Strong Opinions Loosely Held' is the Worst Idea in Tech"
Otherwise it's just double-qualified.
For example: One new tool for me: the pre-mortem. Given a major deliverable or initiative, pretend it’s, say, 90 days from now, and the project is a flaming crater. Go around the room with everyone throwing out ideas of what “went” wrong. Then assign probabilities to the various issues and discuss possible motivations. It’s difficult to overstate how valuable this tool is. It also gamifies what can be considered negative thinking or nay-saying.
Or, get 60% of the benefit by listening to this interview with her: https://overcast.fm/+Ei1D2qbQQ
If that doesn’t grab you within 15 minutes, bail. But if this topic is interesting to you, I suspect you’ll easily devour the whole thing.
(Edited to add pre-mortem)
My perception is that open-mindedness and nuance in discussions have been steadily decreasing, replaced with hyper-confidence and binary thinking, and the rate of that decrease is accelerating. I wonder if anyone other than me has noticed this change in the nature of forum discussions (here and elsewhere?
> Even if someone does have the courage to push back, in practice the original speaker isn’t likely to be holding their opinion as loosely as they think.
"isn't likely" is a massive understatement in my personal experience, not only when the topic of discussion is technology, but even more so when the topic is politics.
> So, what about the situation where someone else goes first and makes an absolute statement? There is a simple ninja move! Just say “It sounds like you are 100% sure of that, is that right?” If the answer is yes, you can ask them to explain why they are certain and see if they have any data to back it up....
This does seem like quite the ninja move, I like it.
> ...If not, you’ll have prompted them to assess their actual level of conviction, sharpen their thinking, and open up the conversation. It is a simple, kind way of helping them develop a style of thinking and communication that will improve your organization.
I am very curious if this ninja move will indeed be received/interpreted in the real world as a "simple, kind way of helping them develop a style of thinking and communication that will improve your organization." I predict (80% certainty) that kind of reception will be quite rare (<50%).
What has changed is YouTube culture, which seems to have a scary number of people SHOUTING and EMOTING in VERY EXTROVERT WAYS DIRECT TO CAMERA with plenty of FAST CUTS and ASIDES, even when making videos about technical content.
Sometimes it's entertaining, but just as often it's tiresome and distracting. I wouldn't object if it was toned down.
I'd bet (there's no way of really knowing obviously) that back then, that worked perfectly fine because those people for the most part actually held their opinions loosely. Oh, they wouldn't give them up without a fight, but what was Correct trumped personal opinion. I don't think there's very much left of this in modern culture, even among those who think they behave this way, at least on certain topics. Distinguishing between reality and one's perception of reality does not come naturally.
Non technical managers are particularly susceptible to this kind of thing. I find a lot of developers who pick up on this act with overconfidence as a result.
They also write like this.
Short sentences, free of uncertainty.
They give direct orders, separated by 3 newline characters.
Beginners are surprised.
If you don't do the same, you cannot count yourself among the best.
(This is satire, but if you're online enough to discover this group, you'll recognize them.)
Yes, I've tried to work with techbros whose thinking is "debate me until one of us gives up, last one standing is the winner and it won't be you". It's not good.
It can be drilled many layers into, or it can simply stated as having this work on your team requires everyone meet a certain level of maturity.
If every team member understands the axiom that the singular reason for anyone's participation in the discussion of the idea/topic is strengthening the idea/collective understanding of the topic and can maintain that as their primary motivator when deciding how to act then you will find that what was interpreted as toxicity before is then interpreted as passion, and that the toxicity previously motivated by fear of looking bad is now reasoned into nonexistence because engaging in reasonable discussion via admitting to a blindspot is in service of the ultimate goal.
You need some combination of things that get you the end result of everyone cognitively keeping their emotional reactions at bay if you want this style of thing to work on your team.
edit: also, IMO the whole adding a percentage certainty qualifier to your statements misses the point, bless its heart. The reason discussion gets better results is because you get more data. Slapping a percentage estimator to your certainty implies that you know about how much data you already have (true) and that you know how much data is all the data (false). No one knows that. This is another axiom that's very helpful in diverting the emotional shrapnel that this style of debate can otherwise toss out because it allows everyone to be comfortably humble. Comfortably so because they will not be humbled by another person, but rather by the undeniable infinity of the universe, which some might call God.
Then they go to ask an overconfident jerk who is maybe 10-20% correct, but expresses it as 100% certain.
Then I get the reputation for being the one who never knows what he is talking about, and the next time something comes up they won't even ask me, they will go straight to the overconfident one who consistently gives bad advice with no room for doubt.
So I guess the short way of saying it: I think it's great to communicate when you are not sure, but you need buy-in from everyone that this is what everybody does. You can't have people that assume uncertainty is weakness, or expect that the world is more certain than it is. And it can get troublesome when you have a mix of different levels of honesty and humility. People get shafted for being honest.
I learned this lesson myself, people interpret uncertainty as simply "not knowing", if you "not know" enough you're considered not knowledgeable. I had thought stating my certainty would mean when I was very certain, people would take that information seriously. It was the exact opposite. Simply being uncertain means you arn't reliable and can't be trusted.
I've mostly abandoned this way of thinking now. I don't even think like that internally any more either. It was not a useful challenge/ideology/creed for me. I take comfort that I'm at least no worse than anyone else.
Some people in corporate politics will walk all over you for allowing such things. But I would call it not being a jerk.
Again, I enjoyed the challenge, I just didn't enjoy the effect it had on me socially, or in my career. I'd rather have friends who believe I'm smart and competent on a good career path than always be right.
[1]: https://www.cia.gov/library/center-for-the-study-of-intellig...
I can say that "strong opinions loosely held" is basically an oxymoron and never works in practice, assuming it actually means anything in the first place. The few places I've seen proudly proclaim this motto have been the most toxic, poorly managed businesses I've seen. That phrase has, to me, come to represent a big, giant glowing neon warning sign to stay far, far away.
I wonder what it would take to start a community where we downvote exaggeration, much like we do jokes on Hacker News?
Oh neat, he's talking about me.
I'm terrible for this. I know it. I struggle with it. And I hope that over time I improve, get better at it. Because I've met people like me, ones who don't know it's a problem, and they're very difficult to work with.
What I've done is to take a step back and try to channel communications into asynchronous channels. I find it very challenging, if not outright impossible (again, this may just be me and my personality talking) to try to explain my thoughts and opinions in real-time in a room, or worse a video conference, where you're forced to make all manner of assumptions about other peoples understandings of what you are saying. When it goes async, you can lay out the foundation and point towards specific points in the sort of nuanced and context-rich environment that technical conversations demand.
On the team that I lead, this looks like Slack, email, Confluence, and Jira. Decisions are distilled down from all of these sources and worked on in shared documents, maintained in a decisions log that people can be pointed towards to get a better understanding of the projects. This documentation is very useful for rooting out incorrect assumptions, inaccuracies, confusion, etc..., in the topics that we talk about. It takes a lot of work, of course, but I think the outcome is better. This team has much more cohesion and a higher degree of collaboration (as in boosting each other up vs. simple cooperation) than the other teams I've been part of.
Life is too short to work with stupid, intellectually lazy people.
Be informed and educated, trust your judgement, and be willing to change and adapt as needed.
Wait, what? Why would the author go so far to generalize something like that based on sex and race?
As for the author, I understood the sentence to refer to a lack of diversity in engineering offices.
Unnecessary, perhaps, but not inaccurate.
It reminds me of George Orwell's classic observation (one of many, let's give credit to those who really deserve it) in The Lion and the Unicorn[1]:
> England is perhaps the only great country whose intellectuals are ashamed of their own nationality. In left-wing circles it is always felt that there is something slightly disgraceful in being an Englishman and that it is a duty to snigger at every English institution, from horse racing to suet puddings. It is a strange fact, but it is unquestionably true that almost any English intellectual would feel more ashamed of standing to attention during ‘God save the King’ than of stealing from a poor box.
No longer is it only an English disease, millenarianism has spread worldwide and become (ostensibly) a distaste for whiteness and masculinity. There was at least some justification in being against nationalism in Orwell's time, none I can see in this irrational racism-and-sexism-as-self-flagellation for this period.