Hacker News Reader

Comment by MattGaiser | Hacker News Reader

1. For things like remote work, it only became a trend because of the pandemic. You would have needed to predict the pandemic to predict that management would permit remote work. It was not really a trend, but rather an emergency measure that morphed into a desirable benefit.

2. This only looks at successful things. We must examine unsuccessful things to see if there is any real predictive ability.

View on HN

Regarding point 2, an old joke comes to mind: "economists have successfully predicted nine of the last five recessions".

I am constantly seeing articles about how a market crash is coming and have been seeing them since 2014.

I think part of this is media enjoying amplifying clickable stories. A lot of people (myself included) just can't help but click about news like "market crash is comming" "housing prices are about to crash" etc. Current media business models sadly are driven by sensational news.

See also: all those revolutionary stories like cancer cure, that happen to only work in mice for now and will probably never work in humans, but we still click on those

I agree. I think The Great Recession of 2007 created an ongoing paranoid that marketing folks capitalize on.

I think it’s that people realize that the underlying issues that caused the financial crisis haven’t actually been fixed, and the Fed has been experimenting with policies that have never been seen before to prop up markets. While it’s worked so far it still feels like there are structural weaknesses that could cause massive problems, and just because things have been good doesn’t guarantee they won’t eventually fail, but timing is notoriously difficult.

"Policies that have never been seen before" seem to describe a good part of the last century, which has certainly had booms and busts, but the novelty has been more of a constant. We just don't have that long a history of modern economics. We've been flying by the seat of our pants the whole time.

Sure, but also, there's a generation of us finance folks who learned how the pipes break. So that when they see something like Evergrande, they go 'well it will probably get bailed out, but if it doesn't.....'

https://www.bloomberg.com/news/articles/2021-08-23/evergrand...

https://www.nytimes.com/2021/08/10/business/economy/china-ev...

crashes happen every ten years or so. Predicting them shouldn't be so hard.

Found the millionaire.

Yes I remember the 2017/18 recession

Every prediction I see is: a crash is coming but we know with 100% certainty that no one can predict when. It might be tomorrow or 3 years from now.

Framed that way, it's less a prediction than a statement of the cyclical nature of economies, but it can still grab headlines!

better formula would be “in next 20 years we will have economic downturn”

Nah, I think you haven't been following the permabears.

A lot of them talk about crap like the price of gold, the use of BTC as a hedge, PE ratios, "technicals", candlesticks, simple moving averages, charts, commodity prices, QE, negative bond rates, near-zero interest rates and the like. Some get political even, talking about presidents or policies they don't like and why they'll cause a crash.

Its pretty easy to weave a story around ill-defined feelings, because none of these things determine the market trajectory. You can be 100% correct in your facts and/or analysis of the present but still make terrible predictions of the future.

The permabears always have a good story. They're really good at making stories for why the crash is going to really happen this year.

They might not be wrong though. Irrationally high valuations are signs for a bubble, but do not help predict when the bubble will burst.

There was a crash last year. The only reason it's not really acknowledged as such is that the recovery was unusually quick.

Crashes don't have to be big and the recovery doesn't have to take years or even decades.

Recovery could also prove to be further papering over serious problems. Time will tell how robust a recovery it truly is.

This scares me. Here in Canada, the government scrambled to put "printed" cash into motion, and inflation is already having real impacts... Particularly on poor people. I worry we're only seeing the tip of the iceberg, so to speak. The same thing happened in the USA, but I suspect our economy here is less resilient to fluctuations.

> The only reason it's not really acknowledged as such

Wait, there are significant sources that don't acknowledge the existence of a deep and sharp pandemic related downturn?

Timing is the >one billion question. Markets aren't predictable because they are dynamic so any measurement and analysis affects the behavior of it's agents.

I have seen articles about an imminent market crash since 2009.

I started paying attention to market crash predictions in 2009.

At least you can't say you were not warned.

Long term near-zero interest rates have fueled unlimited growth for companies. Unlimited growth means no ceiling to stock market prices. A correction will come when the Fed raises interest rates. We started to see it in 2017 but the Fed curtailed. It's inevitable though. But to your point, no one has called it correctly yet.

And then when a market crash happens articles say "we don't know what's happening". Look at finance news from 2008 and early 2020.

And then a year later there are articles saying, "Oh, it turns out that the regulation we created 80 years ago that was specifically designed to prevent exactly this situation was repealed 40 years ago."

And then a year after that there are articles saying, "Why isn't Congress fixing this obvious thing that only makes the very rich richer and always leads to economic disaster?"

And then a year after that, there are articles saying the next crash is coming....

> I am constantly seeing articles about how a market crash is coming and have been seeing them since 2014.

I’ve been seeing them nonstop since the early 80s, and I suspect I would have seen them earlier except that my media consumption before elementary school was fairly constrained.

It's a lot like predicting earthquakes right? You can measure strain and know that you're due, but you still don't know precisely what the trigger will be or when it will happen.

We don't make such jokes about seismologists though. Why not?

I had to break up with my Seismologist girlfriend. She kept pointing out all my faults.

But we do make such jokes about meteorologists.

I suppose it's the matter of media exposure: unlike economists and meteorologists, seismologists aren't constantly on TV telling us bad times are coming soon.

If the meteorologist says it will rain on Saturday and be sunny on Sunday, and in fact they happen in the opposite order, my weekend plans are ruined.

But the seismologist is just saying, "I don't know exactly when but it reeaaally looks like we're going to get a big earthquake someday soon better make sure your building is up to code and you have your emergency supplies ready". And we nod, and check our emergency supplies and upgrade our infrastructure, and when five more years go by without an earthquake, and the seismologist is saying the same thing, nobody says "silly seismologist, predicting twelve of the last three earthquakes!"

But when an economist says a crash might be coming soon, and you'd better make sure you have emergency savings set aside, people do make those jokes.

What's different about these scenarios?

Maybe one difference is with the economy there is a lot more obvious benefit to be had for almost any person to expect (pretend?) there won't be a crash. For earthquakes, challenging the seismologists only benefits the niche group of real estate developers. While the cost of worrying about earthquakes is relatively little ostensibly.

I'd heard some argument that humans inherent tolerance for risk is much lower than it rationally should be today because it was evolved in an environment where we were much more likely to get killed. I think human discourse is constantly grappling with that biased instinct.

It's more like predicting Y2K. There was no countrywide carastrophe from Y2K, therefore the computer scientsists and engineers who predicted it were wrong. And by extension, computer science and engineering are bogus. Right?

Seismologists don't try to imply that there's "news" every 6 hours worth reading.

We are accustomed to their predictions being shaky.

As someone who has worked on exclusively remote teams for around 7 years, the pandemic accelerated the trend. It was already very much headed in that direction.

Why is your work situation relevant? If you meant it as a "this is what my bias is" then it further weakens your argument, which may or may not be desirable for you.

Because remote work isn’t new and was on an uptrend before the pandemic.

My dad likes to remind me that he was 'working remote' (had a dial up modem and two factor auth in to whatever he was doing at AT&T/Ameritech) back in the early 2000s

I started working remotely over DSL at the end of 2001.

I also recall getting an ISDN line in the late 90s, but I was working in an office at the time.

I started working remotely end 1994. An analog line ran to my office had two analog 19k2 modems on it and I worked remotely on the server which was connected to the internet using a 10mbit fiber.

Worked from home before that '90/'91 but not online. Had an IBM cash register in my bedroom for which I wrote a POS application.

“It exists and is growing” and it being a trend are very different. I think it is an irreversible trend now but it wasn’t before the pandemic. There are just way more people like myself didn’t even think of it seriously for many reasons (I did know it existed) and they all have woken up now.

Remote work seemed to be declining in the Bay Area pre-pandemic. My company and many others were pushing for more in-office work over remote work as they felt that employees were able to be more creative and productive if they were able to collaborate spontaneously which in their eyes could only happen if everyone was in the office. This really seemed to take off in the area after Yahoo had implemented their no remote work policy.

if course it declines in the Bay Area, a place that people famously move into for work. It'd be silly to move to the Bay only to then work in a shitty rental appartment.

Growing is most certainly a trend. It doesn't make it a permanent trend but that is definitely a trend.

Sometimes people share their bias in the interest of just being transparent. This is mostly because not every discussion is necessarily adversarial, and having more contest is helpful for their companions. However it also makes them appear more trustworthy, adding informal weight to their subsequent statements.

Point 2 notwithstanding, I wonder what's really the threshold of "predicting" for point 1. I've been reading and participating in conversations about remote work on HN for many years before the pandemic; hell, HN is how I discovered this concept in the first place, and what helped me become a remote worker some 3 years before COVID-19 hit the news. REMOTE tag in the monthly "Who is hiring?" / "Who wants to be hired?" threads has also been established before the pandemic.

To me predictive power must be actionable. Is there something in 2018 that would have made you think "I have got to buy Zoom stock"?

That's a good heuristic, thanks.

Ironically, not only nothing made me think "I have got to buy Zoom stock" in 2018, IIRC around 2019 there were plenty of reasons to short Zoom. I wonder if they'd have survived if not for the pandemic.

Before the pandemic, zoom was a starting to gain market share over a bunch of the legacy conference software vendors, like cisco webex. They had their act together on the enterprise sales side and were getting traction via their freemium model ... just dropbox/box got enterprise traction even when companies already had things like SMB/CIFS or sharepoint. The UX was better for the same reasons many an HN didn't understand what advantages Slack has over IRC.

As someone who has been remote, hybrid remote, and almost every other version distributed teams and remote/asynchronous working over the past 13 years, I assure you there was and still is a remote working trend well before COVID. COVID only poured gasoline on an otherwise very long-term "how we work" trend. COVID probably condensed 10 years of where remote work probably would have ended up in the span of 18 months.

To point 2., This article seems like a great example of selection bias. Hacker news discusses many new ideas in the tech world. I'd also like to see things HN didn't like and see the success rate on those services (the famous Dropbox comment[0] comes to mind)

[0] https://news.ycombinator.com/item?id=9224

But Dropbox is already losing marketshare, as OS vendors now implement their own cloud solutions. So you could say that HN was right in the long run.

> You would have needed to predict the pandemic

It's rather short of that of course, but after hearing about an outbreak of flu in China on Canadian television in a hotel lobby on holiday from the UK, very early January 2020, I think the 2nd [*] - I heard approximately nothing about except via HN, which included 'this will be bad/global', until March.

[*] (For a while I remembered, because it was notably earlier than UK newspapers kept claiming as a start date, or 'detected in China', in their charts from March. I didn't claim to know the correct date, but obviously that was an upperbound!)

> It was not really a trend, but rather an emergency measure that morphed into a desirable benefit.

Do you have data to support this? Anecdotally (from own experience and tech workers I know) I feel that remote work has been growing steadily over the past decade, with a an increase in job ads mentioning remote (even if they went out their way to say "office only") but I don't have data either way to back this up. It feels that long-term changes in tech as well as management culture were making remote more common in the long term, and the pandemic accelerated a trend (much like the shift in the movie industry from cinemas to home streaming).

Regarding point 2, isn't it pretty common ( or at least was ) for comments to disparage potential projects?

For example, the infamous Dropbox comment: https://news.ycombinator.com/item?id=9224

Pandemics have worked as an accelerator or even as a catalyst for technological change since at least the Black Death and this one certainly is no exception.

However, remote work didn't become a trend because of the pandemic, it became the default scenario for many jobs that don't actually require on-site presence but that due to cultural inertia nevertheless often still happened in a colocated office pre-2020.

Remote work has been a trend since long before 2020. With the pandemic the proposition in many cases now simply was "Either work remotely or see your company go out of business.", which had many companies reevaluate their ways rather quickly.

> You would have needed to predict the pandemic

That was not difficult to do, even in Late December. Most discussions were taking it really casually, and believed that none of that would happen, which is exactly why it happened and how I predicted it. People are be filled with unfounded optimism, overestimate their skills, and underestimate their bad lack.

Combining the three with the belief that most humans are incapable of understanding exponential and autoregressive processes, it became clear.

But all of this is anecdotal evidence and could very well be an instance of a bear predicting a depression, after all, bears predicted 10 out of the last 3 economic crises.

Not sure if it is true in general. But from the perspective of the software industry, our company has been remote-first since its inception. We do encourage coming to the offices, but it has never been mandatory. I think there is a fine balance in this. Personal time saved vs. the social aspect. And of course some problems just get solved faster in front of a whiteboard.

This would have been a better analysis if it ignored headlines and used comments, perhaps in combination with some sentiment analysis. Also, nfts, Tesla, etc aren’t really tech trends but rather consumer and valuation trends. They should have tested for specific technologies like programming languages, paradigms, frameworks, and libraries. Or specific hard tech

Would you get the same result if HN is a popular content marketing forum that’s frequently used to promote new businesses and products? Also, I skimmed the graphs and only concluded that HN contributors have a short attention span. I must lack the sophistication to interpret the results.

for 1.. I'd say it was a trend...just a much smaller one.

I've been 100% WFH in tech for 20 years, and it got easier every year even before covid. What you had to predict, wasn't the pandemic but what tech would be hot enough employers wouldn't care where you worked.

I'm pretty sure overlaying Google Trends onto keywords in HN is more of an informal analysis than a rigorous decision process. True predictive value needs a more rigorous model, but these sorts of data explorations are a great way to motivate folks to dig deeper.

For things like remote work, it only became a trend because of the pandemic

I think it was more of a catalyst.

Remote work was definitely a trend, the pandemic just accelerated it.

Edit: for those downvoting, here’s the data:

https://slidemodel.com/remote-work-key-trends/

I was surprised to see you downvoted. Did people forget that "digital nomad" was trending years prior to the pandemic?

> Did people forget that "digital nomad" was trending years prior to the pandemic?

Often encouraged by people selling tools, books, coaching, etc. related to the digital nomad lifestyle.

And often are self-employed/freelance

I always interpreted digital nomad to be a Tim Ferris type person working 4 hours a week and living off relatively passive income or some kind of online self employment, not employees.

It was definitely a trend, fully remote companies were coming into vogue long before the epidemic.

Thank you for this.

I am actually surprised as it seemed that with IBM, Yahoo, and various other companies dragging everyone back to the office in 2017-2019 that remote work was in retreat.

3. Hacker News comments have been vehemently anti-cryptocurrency for a decade. I remember being shocked how out of touch with actual hackers this place is.

Reminds me of punks who'd shop at Hot Topic.

1. We still don’t have any proof crypto isn’t just a huge ponzi-like scheme. In fact a lot of them have been shown to basically be just that.

2. There is nothing wrong with “punks” shopping at hot topic, whatever you mean by punk.

> 1. We still don’t have any proof crypto isn’t just a huge ponzi-like scheme. In fact a lot of them have been shown to basically be just that.

What would that proof need to look like?

A successful venture using cryptocurrencies for anything other than speculation would be a good start.

What about https://www.valiu.com/en?

It's just another speculation scheme. "Protect your money from inflation" means "make a bet that our token's price will go up".

If you want to safeguard your savings against inflation, the way to do it is to invest it in vehicles that are backed by something, like corporate stocks, government bonds, real estate, or commodity money, not a vacuous line on a digital ledger supported by nothing but people's hope that it'll be worth more tomorrow.

Valiu's claim that they're backed by USD is dubious at best. Tether was running a similar scheme, we now know that they only have 2.9% of the money that they claim to have in reserve, the rest having been traded away to mysterious LLCs in the Caymans, and many of their coins printed out of thin air.

1. Its continued success is proof enough, as people conduct legitimate business with it.

Separate BTC/ETH/XMR/etc from random shitcoins and NFTs.

The only "legitimate business" anybody's done with crypto is buying lots of it and hoping the value keeps going up. Nobody is actually using BTC/ETH/XMR/etc to do business or conduct commerce, it's all just speculation.

There is no compelling reason to believe that cryptocurrencies are anything other than digital beanie-babies. At least beanie-babies had inherent value as a toy.

> The only "legitimate business" anybody's done with crypto is buying lots of it and hoping the value keeps going up.

Yes, and HN kept telling everyone for years not to do that. And the value kept going up.

I wish I had a bitcoin for every "don't buy bitcoin" article that hit HN frontpage...

Beannie baby prices were pretty high in '99 too.

Speculation is not business, high prices are not proof of utility, and the purpose of currency is not to hodl.

Actual punks or just punks?

"actual hackers" use cryptocurrencies because they are easy to steal and launder. Despite the name, most of the people on Hacker News are not interested in the criminal part of "hacking".

Reply on news.ycombinator.com