Cut out everything that's not surprising (2019) | Hacker News Reader

Cut out everything that's not surprising (2019)

sive.rs

110 pointssurprisetalk59 comments

> People only really learn when they’re surprised.

This is a nice sensationalist punchline, but I don't think it's true at all. Most of what people learn is not bang, a new thing unlocked, but rather repetitions and small incremental improvements to truly hone and understand an idea or activity. Think of learning a language for example. You wouldn't say you've only ever learned something when you hear a new word for the first time. The real learning is remembering it, using it in different contexts, understanding how other people use it, seeing it in slang, etc.

I agree repetition is key. But here's what I noticed from personal experience. When I first read something interesting, it might not stick.

When reading it again (say: the same book a year later), knowing I must have read this very interesting thing before, but apparently forgot I more or less have the same reaction as being SURPRISED. And more often than not THAT is when it really sticks.

This is from anecdotal, personal experience, but being surprised sort of unlocks things in the brain that make things stick (I could go on, on how this would be a biological advantage etc. but that is reaching) and I guess that is what Sivers is pointing at?

I read somewhere that it takes ten exposures to a new word before you learn it. After reading that, I started tracking how many times I see a new word before I recall what it means. Unsurprisingly, it falls along a distribution, but for harder words it does take me quite a few exposures before learning it. By the time I learn it, the feeling is often recognition that I have trouble with this word, and boredom that it's this word again. =)

“I cannot remember the books I've read any more than the meals I have eaten; even so, they have made me.”

- Ralph Waldo Emerson

> This is a nice sensationalist punchline, but I don't think it's true at all. Most of what people learn is not bang, a new thing unlocked, but rather repetitions and small incremental improvements...

The article is specifically about writing talks, or possibly articles. In that context, using surprise to ensure your idea lands makes sense.

> The article is specifically about writing talks, or possibly articles. In that context, using surprise to ensure your idea lands makes sense.

The article literally says "Use this rule in all your public writing. "

It's garbage article.

> This is a nice sensationalist punchline, but I don't think it's true at all

It's only not true because the qualified it with "only". But "surprise" and "shock" are some of the heuristics the brain uses to prioritize the retention of knowledge. "Competitive memorizers" (people that compete in memorizing things) often use techniques of associating dull information with shocking mental images that help with knowledge recall.

In my experience it is partially true.

Topics punctuated with surprise will help people learn the material presented.

Yeah, if you take it out of context, it's not an universal rule.

People don't get small improvements or practice from reading an article. The advice is spot on in the context it's presented.

I think this is a mistake unless your audience already knows and trusts you. The world is full of people saying surprising things, and almost all of them are making it up, either accidentally or on purpose.

The people who discover surprising things are usually akin to Darwin and Einstein, people who master current, conventional knowledge well enough to spot new patterns (Darwin) or find creative, productive ways to stress the old knowledge (Einstein.)

You don't have to be on the level of Darwin or Einstein to discover something new, but you should show your homework enough to convince people that your surprising take might be grounded in knowledge and not in ignorance. You will naturally cover this ground in the process of establishing context for your discovery, and in showing that your new observation is poorly explained by existing conventional wisdom. Doing this work is the minimum due diligence to build good faith confidence that you're raising something new and not just learning in public, so you might as well present it to your audience to establish your background and your bona fides.

(Learning in public is fine, as long as it's presented as such. "If you're a beginner with X like I am, you may or may not have discovered" is a fine way to start.)

I think this unreasonably splits the world into cutting edge research, and learning in public for beginners.

In reality, almost every talk you're going to see would be considered educational, somebody mostly saying things other have said before, maybe in a new way, to an audience who is new to that topic.

Also the world is really highly dimensional and there are a lot of things to learn. "Let's talk about sorting algorithms in python" "let's talk about what makes a superconductor" let's talk about how to make your own kombucha at home".

If every talk began with an infantalizing "we're both beginners because I didn't invent this topic" we'd all get sick of it real quick.

I don't think the original article makes sense in the context of an educational talk, so I didn't read it that way.

For example, if you're talking about sorting algorithms in Python, different readers will naturally find different things surprising, depending on their background (Quicksort? Timsort? neither?) and even if you wanted to target a particular very narrow audience that is supposed to be in the same place in the learning process, like in a university course, it would still be bad pedagogy to jump straight to the new knowledge without connecting to something the audience knew first. I think the advice in the article applies so badly to that situation that it would be unfair to think that the author intended it to be read that way.

The article makes a lot more sense if you apply it to speaking/writing for a professional audience with a shared set of knowledge or conventional wisdom, and adding something that you think adds to the shared body of knowledge or contradicts some element of it.

> "If you're a beginner with X like I am, you may or may not have discovered"

What’s a beginner?

Reading 3 books on $topic will give you a deeper understanding and appreciation for $topic than 99.9% of the population. Yet you are still at basically zero compared to the people who wrote those books.

The people who wrote those books are at basically zero compared to the bleeding edge researchers who invested entire lifetimes into 1% of what the book covers. Those researchers are at like 60% compared to the researchers who are looking into the other 99% of what the book covers.

By and large we are all beginners in almost everything.

So be more specific. Share your background. The important thing is to give your readers or listeners a fair chance to judge whether what you have to say will be interesting to them, because if you don't, they're going to default to ignoring you.

I had a math professor in college, whom I had for Real Analysis. He told us that one thing that helped him learn was to ask somewhat-obvious sounding questions to try to make connections to things to check understanding. There are at least two good reasons for this: (1) if you don't understand the basic (unsurprising) things, you probably won't understand the more nuanced things, and (2) what counts as surprising varies with the audience.

If someone were to come to me looking for advice along these lines, I'd say: sure, focus on the surprising thing, but it has to be grounded in the familiar, and what counts as "familiar" depends on the audience.

Precisely. I think the author is saying a talk should have a thesis statement that is novel. That is true.

But to convince the audience of the thesis statement, the speaker needs to cite evidence. The evidence cannot be surprising.

Since most of the talk needs to be evidence, and therefore not surprising, it’s foolhardy to eliminate all that is not surprising.

When people are surprised they don’t put the information through their usual bullshit filters.

So people are willing to hear stuff that’s possibly utter garbage when it’s new and surprising.

The brain has to be open to novelty without dismissing it as untrue based on previous experiences.

So you can say things like “cut out everything that’s not surprising”, and because it’s new and surprising, your brain gives it a free pass: “hey that’s a new concept, could be good, let’s not put that through the bullshit filters just yet”.

A little more careful thought might reveal that the speaker is talking a pile of rubbish.

This explains why people love to go to conferences and rah rah events because they’re going to be fed new concepts they can latch onto without critical thinking.

> When people are surprised they don’t put the information through their usual bullshit filters.

Huh. When I see information that's surprising, that's when my critical analysis and bullshit filters naturally spring into life.

Something surprising is something I either hadn't thought about before, or had thought about but come to a different conclusion with. Something surprising is ripe to be considered from a bunch of different angles. How does it fit in with the things I already know? It is plausible? Does it actually contradict my current model of the world? Might it complement my current model of the world?

Anything that isn't surprising is what gets the free pass on my bullshit filters. Yay, confirmation bias!

Yes, but surprising to whom? If you are going to talk about it, then it's no more surprising to you. How do you know if it is going to be surprising to your audience? When everyone has access to Youtube, TED talks, Social Media and blogs, it is hard to gauge what is surprising and what is not anymore.

Write for an audience, read the other work that audience reads, that will help you get a sense of whether something is going to be surprising to them.

Here is a longer lecture on a similar idea from U Chicago Writing Program https://www.youtube.com/watch?v=vtIzMaLkCaM&t=3s

Indeed! Plus, whether something is surprising depends on what was expected in the first place. So you have to create some shared expectations first, right? Some kind of buildup?

Seems like a spectrum, if you stay too close to the familiar, you will be boring. But if you deviate too far away from the familiar, you will be confusing. A nice surprise lives in a sweet spot somewhere in the middle. Like the very narrow band of our planet's orbital radius where we are close enough to the sun to keep water liquid but not so close that proteins would disintegrate.

From an information theoretical point of view, they are just saying to not transmit non-information :)

Anyway: https://xkcd.com/1053/

You can be interesting without being "surprising"

> People only really learn when they’re surprised. If they’re not surprised, then what you told them just fits in with what they already know. No minds were changed. No new perspective. Just more information.

To me, knowing "more information" seems to be essentially the definition of learning.

Structural learning is different from factual learning. Consider "World War II started in 1939 when Germany invaded Poland" vs "Adolf Hitler rose to power with the intent of exterminating the Jewish population". Memorizing the dates and countries involved doesn't really teach you anything--it's just another war over territory. But hearing about the people and purpose might very well alter your entire view of what humanity is capable.

To a shallow extent, both are "more information", but as per the OP, "No new perspective. Just more information" implies that something beyond simple information is essential to "real" learning. If your perspective doesn't shift, what's the point?

For coders, this is like the difference between learning a new syntax vs learning a new paradigm. You can learn a dozen languages but if they're all just skins over ALGOL then you haven't learned much. But if you know C and you learn Lisp or Forth or APL, it may even change the way you write C.

Overall I agree with him (I don't want to consume information that doesn't surprise me in some way, and technically speaking, it isn't even information without that surprise). But ... "surprise" is totally contingent on audience.

A lot of things that will seem like unnecessary throat-clearing to experts will be both surprising and totally necessary for people lacking context. If you can't walk your audience into the proper context, whatever other surprise you offer will be meaningless, and they'll just tune out.

I think people here underselling this idea a bit. I would like to see this concept played to the death: Where some lecture on a hard subject is structured such that it would focus on the surprising things first and focuses on them. Sometimes I would listen to something complex and I cannot make heads or tails of it. Maybe structuring lessons by how they are surprising, bringing attention to them, would help to clear the fog quicker.

Our attention spans have been ruined by the modern age of the internet! If an audience needs to hold a train of thought for more than three minutes, they're gone. Surprise, anecdotes, humor, problem statements, etc... are great tools that stoke the fire of focus.

Another issue is the amount of attention you get to work with. Each surprise moment spurs an increased attention which decays. So, one introduction to the problem is not enough, you have to break things down! Introduce a complication on the original problem, then immediately fix that.

To each their own, but thinking with this pattern for my talks has worked really well for me. Even for mixed audiences (engineers + PMs + management), everyone gets something out of it, and finds their own questions to ask as well. Makes it easier for them to connect it all to their own problems as well.

This makes me think of the critiques of Malcom Gladwell - essentially, that by focusing entirely on counterintuitive / surprising results, he often misrepresents the concepts he discusses...

> These ideas are far from obvious, self-evident, or trivial. They do have the property of triggering a pleasurable rush of counterintuition, engaging a hindsight bias, and seeming correct once you have learned about them. [...] I think the result is the propagation of a lot of wrong beliefs among a vast audience of influential people.

https://slate.com/technology/2013/10/malcolm-gladwell-critiq...

I'm not sure I would use "surprising" as the indicator of what should stay or what should be cut, but for sure there should be a goal of removing all which is not essential. This is especially good when you provide footnotes or links to additional information for those wishing to dive deeper.

Way back in school I recall the difficulty of writing an essay within a maximum number of words. Inevitably I would be way over the limit and be forced to edit with an eye for brevity. And to be fair, when being brief, content which is surprising to the consumer perhaps activates their brain more, allowing them to fill in gaps better.

Oddly, the author doesn't follow their own advice. Nor do they support the claim on which the entire article depends:

> People only really learn when they’re surprised.

The idea that people learn when surprised is not surprising. The idea that people only learn when surprised could be surprising, but only if substantiated. I would be surprised because I can think of numerous times that I learned when not surprised.

Surprise is fine as a technique to draw the reader in, but it puts you on the hook to support the surprising claim. And the act of whittling that claim down to something supportable could very well leave you with nothing very surprising at all.

While I agree that you should definitely have something 'surprising' in things you write or speak about; I definitely disagree that you should try to cut out everything but the surprising stuff.

Every surprise should be surrounded by supporting information that most people are already familiar with. If you don't include some context that people already understand then the surprises will seldom gain real traction because people will just dismiss it as BS. The world is full of stuff just made up for shock value.

I'm not sure if its worse to have too much or too little in your writing. I've read books where I skip whole chapters because nothing important happens.

I assume on the other end there would be books with missing pieces where important information is not mentioned or explained. I haven't read a book like that (yet) and I think it's because most editors would not allow an unfinished story to be published.

> Nobody has ever complained that a talk was too short.

People did complain. In some of my writings, I regret of not expounding a topic a bit more.

That's a terrible advice. If you write in a way where writer is constantly surprised and not just interested you're a terrible writer, at least when it comes to passing knowledge, vs making sensational articles for ad revenue

In my experience, people learn from struggling to do something. Surprise is only one of the emotions involved. I would say the frustration and dopamine cycle is the biggest factor pushing a learner to master a subject.

You can prepare your audience for a surprise. You need to draw a surprising result from your previous statements. I guess this could be called a rhetorical surprise.

This seems like a great way to make people believe you're smart when you're really just constructing "surprising" talks.

Both surprising news and unsurprising news are valuable, you just update your confidence in certain beliefs in opposite directions.

Even if we accept the premise, the problem is that what's surprising varies per person.

Particularly if you are an expert on the topic you are presenting, chances are that stuff that seems obvious to you will be surprising to others.

If you give a talk focusing on the limits of your knowledge (as a proxy for "surprising stuff"), most likely you'll lose most of your audience.

As a corollary most good talks will have a big chunk of stuff that will seem obvious to you... because it won't be for a lot of people.

Relevant XKCD https://xkcd.com/1053/

from the comments section: Don't audiences sometimes want someone to say something that fits and affirms their views?

They want a mix of both, I'd say.

People who go see a populist politician give a speech expect mostly to hear things that will reaffirm their existing views. But the rousing and titillating part is the blended-in surprise that explicitly goes against what they used to believe — e.g. "To preserve our beautiful democracy against these ruthless internal enemies, we must now [do thing that's directly against democracy]"

> People listen to a talk, or read an article, because they want to learn something new.

That’s surprising. /s

This is opposite to how LLMs select word choices.

This take is unsupported by pedagogical research

sive.rs is not surprising, so...

> People only really learn when they’re surprised. If they’re not surprised, then what you told them just fits in with what they already know.

And under that is picture of quad-bike, where wheels fell off!

People do learn from mistakes. If you are "surprising", and loose all wheels, people will learn from that mistake! But they will never ever interact with you again! You will get on banlist as a moron and clown!

Real value is in boring, predictable and repeatable stuff that delivers. Like I spend 90 minutes on lesson, and it gives me 1.5% skill increase. I do it 100x times and I have something. You will never be able to "surprise" me and deliver value 100x times! Smart people do not eat junk food!

People learn in different ways. I remember details of 9/11 because they were surprising not because I read about them 100 times.

That's not knowledge, that's trivia

If you say so.

You won’t be able to change my mind with an empty quip - people can and do learn by being surprised; and other folks learn differently. I’m not sure there’s much room for debate about that, but if it makes you feel good, spend the energy.

In this context, learning something new is surprising.

> People do learn from mistakes.

Mistakes are surprises.

> Like I spend 90 minutes on lesson

Assuming this lesson is teaching you something you didn't know, that's the surprise.

> Assuming this lesson is teaching you something you didn't know, that's the surprise.

There was some trick in calculus that makes it 1% more efficveint. Big deal but hardly surprise.

People on HN have real interpretation & ego issues, it is unbelieveable.

Everyone seems to be discussing whether the author is correct about claiming that "people only really learn when surprised", he wrote it that way to give emphasis to what he had previously written, not really used it as a fact.

So this comment section became a battle of proving the author wrong, when in reality, he has actually written a very interesting piece with a nice hook. I don't understand why people nowadays can't just take the good part, the lesson learned and move on.

It seems like HN has been going downhill since years and it only gets worse. Sad. Well, that isn't a surprising message right? Maybe I should have omitted it.

Yeah ... I get the sense people are ripping that sentence out of context and debating it on its own merits, when there should really be no debate because, taken literally and context-free, the sentence obviously doesn't make sense.

It all seems like such a waste of energy when we could be discussing the substance of the article.

It's a very short article. It's not like it's full of substance and that one sentence is a faux-pas or overenthusiastic. The content is basically that one idea plus some padding.

I read it whole. It's terrible and makes some weird claims.

It's like author didn't understood difference between "surprised" and just plainly interested.

Reply on news.ycombinator.com