The Evolution of Hacker News (2013)

I've been reading some old (olde?) tech references, forecasts, and futurist works of late, and am noting a few trends.

Among them are Alvin Toffler's Future Shock, which turned 50 last year, read for the first time, Lawrence Lessig's Code and Other Laws of Cyberspace (22 years), a re-read, and Andrew Shapiro's The Control Revolution, also for the first time.

All have hits and misses. Some of them are more slanted one way than the other.

Toffler has aged surprisingly well, in my view, despite some weaknesses. Lessig, at least in his introduction and first chapter, similarly. Shapiro is rapidly shaping up to be an excellent Really Bad Example. (Lessig references Shapiro early in Code, hence my looking into it.)

I've been toying with a notion of an ontology of technological mechanisms, which looks not at fields of technology, but rather how they achieve their effects. I've come up with nine basic categories: materials, fuels, process knowledge (what's generally meant by "technology"), causal knowlege (roughly, "science"), power transmission and transformation (ask if you're curious), networks, systems, information (accessing, storage/retrieval, processing, disseminating), and a final category I've tended to call hygiene functions -- dealing with unwanted or unintended consequences.

It's that last aspect which seems to dominate considerations ultimately, because all technologies can be thought of as interventions in some system to an intended effect, but having several dimensions, including:

- Benefit / harm

- Near term / long term

- Clearly evident / non-evident

Generally, we tend to choose technologies with clearly evident near-term benefit, and strictly avoid those with clearly evident near-term harm. In cases where a mixed set of benefits and harms of varying evidence or perceptibility, and of differing timeframes is present ... things get more complicated.

And as interventions in systems become more complex, the odds of a negative interaction tend to increase.

The upshot is that cautions are often more significant than ethusiasms, and in the case most especially of Toffler, the concerns he raises, most especially of psychological and sociological impacts of increasing information flows and rates of change, do seem to have been reasonably prophetic. Sections dealing with specific technologies and their presumed social benefits are the weakest. In some cases the promised benefits have come to pass, occasionally to such a degree that it's hard to even see them from the present vantage point --- they've very much become part of our world in a way that is like water to a fish: so ubiquitous it's easy to forget it exists at all. In particular, the developers or advocates of a specific technique (or occasionally, scientific advance) seem to be exceedingly poor at conceiving of, or at least sharing any conceptions of, downsides. There's an extraordinarily strong positivity bias. Not universally, but as a general rule.

Lessig is similarly concerned with harms, and (at least in its introduction) seems to focus accurately on the right problem areas and dynamics.

Shapiro has read to me through the first part of his book as, charitably, remarkably oracular in the sense of "this is a prophecy which might be read two ways", as in "if King Croesus crosses the Halys River, a great empire will be destroyed." (Spoiler: one was. It wasn't the one Croesus had in mind.) For the most part, Shapiro reads to me as childishly naive, credulous, fatuistic, uninsightful, concerned with the trivial, self-parodying, and often foreshadowing but apparently with absolutely no self-awareness in doing so. It is remarkable how many of the specific actors and situations what are front-of-mind today are mentioned or alluded to in the text. But the references don't seem aware of their own significance.

In his defence, Shapiro occasionally evidences some degree of awareness or perception, though these feel like brief moments of lucidity in a once sound mind. And it's possible that the latter parts of the book refute the naivete of the first chapters, though I'm skeptical (and reviews I've read suggest otherwise). It's very much a case of The Author To Whom I Must Constantly Scream as I read the text though.

But yeah: be very skeptical of self-involved proponents of technological or other initiatives. Whatever expertise they may have is strongly moderated by deliberate or unconscious self-serving bias.

I'm not convinced there's an actual problem with downvotes on HN. It mostly seems that way when you take a snapshot of a page, but waiting an hour (if the comment section is active and the topic is not controversial), grey comments have a tendency to become un-grey over time unless they fall into a few categories. Maybe someone doing NLP research and sentiment analysis could put together a more thorough analysis, but from what I've seen:

Comments that tend to become and remain grey:

- Complaints about downvoting (generic complaints like yours, or specific complaints about an instance of downvoting)

- Complaints about paywalls

- Complaints about the article based only on the first couple of paragraphs

- Complaints about article titles

- Other, generally off-topic, ranting comments

(NB: Many of the above fall under the broader category of "boring and off-topic")

Comments that tend to become and remain dead:

- Bigoted comments (rare here, but happen, see a discussion earlier where someone suggested non-white people should be gassed)

- Highly political comments

- Comments consisting mostly of personal insults directed at the previous commenter or article author

Discussion topics where, to the extent that it does occur, "downvote abuse" is most apparent:

- Anything political

- Anything dealing with race/ethnicity/national heritage

- Many dealing with US immigration policies (which hits both of the above)

- Many on economics, especially where there's no "one true answer" and opinion rules the day (same issue political discussions often have)

Interesting thoughts, thanks.

> - Benefit / harm > - Near term / long term > - Clearly evident / non-evident

I still think that in addition to these things it's the growth function, the linear segments of it, then sudden non-linearity that we fundamentally always get wrong in our predictions.

Think of the 1960-1970s that was considered the era of space exploration or the beginning of it, and how people wrongly extrapolated it up to the year 2000 (A Space Odyssey, yep).

Or when we try to predict non-linearities, we get it wrong too. Taleb predicted the end of Google in 2007, also made fun of e-scooters back then. But if I'm not mistaken he also praised segways. Well, everybody did.

In other words, we are wrong most of the time and we are wrong about the non-linearities in the growth functions. So much so that you could probably even say: if a prediction is a linear extrapolation then it's definitely wrong. If it's a prediction of a breakage then the chances of it being correct are purely random.

Look at all the predictions of the next market crash made in the past 10 years. It's as if the market is listening and doing the opposite!

> I miss PG being on, but I completely understand why he left.

pg isn't shy of (seemingly) speaking his mind on twitter though.

I disagree. Semi-early HN (I've only been here since 2010) was a lot more political and a lot more technical. When the current heavy moderation regime was added, political topics were intentionally (and for a period openly and explicitly) pushed out. It was after a period with a lot of threads about black people's experiences in tech.

I thought the moderation would chase me away, especially with the reduction in technical content (after a long period of X in javascript threads overwhelming the site.) I was wrong, it's still nice, it's just more mainstream. It's the same stories that are trending on google news, just debated and discussed by smart people with grammar and civility fetishes. That's not actually bad.