Hacker News new | past | comments | ask | show | jobs | submit login
Ask HN: What were your tech predictions 10 years ago?
13 points by ipnon 10 months ago | hide | past | favorite | 24 comments
There has been such a flurry of technological progress in the last few years, it seems to me this is a good moment for reflection. What did you think 2024 was going to be like 2014? What did the future look like to you then?

Personally, I distinctly recall this feeling that Facebook would be the largest and greatest tech company in the world for some reason. That obviously has not turned out to be the case, no offense to our many esteemed friends at Meta. Clearly the popular culture of the 2010s was not capable of predicting the future of our industry with much accuracy, and social media networks seem to have sufficiently saturated the world now.

My second distinct recollection is my belief that we were in the midst of yet another AI winter. Although this seems patently false in retrospect, as many things do, I remember this was not a fringe opinion on even HN at the end of the decade (now it is clear to me that the pioneers in this emerging field were busy hacking instead of posting). It was thought that attention and the transformer were the great breakthroughs, and that we would have some marginal improvements to Google Translate and other such applications that seem quite humble in comparison to what is now being released. To claim that advancements would be made so rapidly that the goalposts on AGI would shift on a daily basis would have been preposterous!

I don't intend to rant too much but it is remarkable how as a young programmer and computer science student it was really quite difficult to foresee how unpredictable and varied the development and use of my art would be. I only humbly suggest that we once again be cautious about the obviousness of our current predictions for 2034.




I expected crypto to actually be adopted by a larger market. Since then I've been disillusioned and I have completely lost faith in it. It has been 15 years of Bitcoin and the usage is still pretty much limited to speculation and online scams. I can't think about a single popular online store that accepts any cryptocurrency.

On top of that the price of BTC is so disjointed from reality that a singular transaction at the time of writing this post costs ~12$. Imagine taking a single trip to a grocery store, a bakery and a meat shop and paying 36$ for, well, paying. It's insane.


Well, as an economist by training, I knew economy works thanks to trust. You can't buy anything without trust even in a (real) marketplace by swapping cash and the commodity. There's still a enough room for fraud.

But tech industry doesn't listen to humanities and tried it themselves to make a tech that assumes no trust at all. Well, it actually has been a good experiment -- now we know what's the cost of total distrust in a perfect model market.

And this extends to other areas. For example, in my country, road police service made a system where you'd book an exam and come alone, and just do it, to eliminate local gatekeepers. Well, the gatekeepers (local road police and driving schools, that consist of ex-policemen) just do it the same way, but ask you to log into that system and book an exam at any time those days, but come and do what they. IDK what in particular the government tried to fight, but people can both fool and fulfill these procedures -- only a real human could possibly notice the violation and change how things work.


I think the main thing I got completely wrong was how tech would impact people's lives.

I was so optimistic 10-15 years ago. I thought the internet was genuinely going to connect to the world. I thought the world would grow so small that national conflicts and rivalries would seem silly.

I also thought people would want freedom. It really shocked me when people started to be deplatformed around 2016. But more than the deplatforming itself the silence and lack of outrage was the thing that shocked me the most.

I think really the only thing I kinda got right was the growth of AI. Around 2012 through 2015 there were decent number of AI breakthroughs happening. I had a bet around this time with a friend who was a machine vision researcher that by 2030 people would have started losing their jobs on mass to AI. I may still be a little early on that, but right now we're definitely ahead of my expectations in many ways


I was dismayed at the lack of Capability Based Security, and hopeful it would be widely used by now[1], making computer security largely solved problem. Instead, most people believe one of two falsehoods: it's a matter of hiring smart enough people, or that it can't be solved at all.

I was really excited about the idea of self driving cars.[2]

[1] https://mikewarot.blogspot.com/2014/01/its-nearly-impossible...

[2] https://mikewarot.blogspot.com/2013/01/self-valeting-cars.ht...


We would not have self driving cars or any other useful robots in quantity. China would wall itself off from the western parts of the internet to an even greater degree. Blockchain will never find widespread use.

I thought face recognition tech would be more widespread, but it turns out that its type 1 and type 2 error rates are far too high for many applications. I also believed that what Theranos was selling was feasible.

I'm more impressed by what hasn't happened than what has.


In 2014 I was thinking: why don't we all (developers) work remotely? I hate the commute time.

Nowadays I do work remotely. Thanks Covid!


I expected more work on distributed computing and data. Data that lived replicated on systems all over. So the internet wasn't so centralised. Less "amazon owns everything". I thought the block chain and bitcoin would push people this way.

I was very wrong. We moved everything to the cloud and amazon is the cloud.

I dreaded the idea of "serverless" and micro-services, and I'm happy to say for the most part those didn't catch on. Or where they did, they've begun to putter and die.

Predictions for 2034 - Less internet freedom. More tech to ferret out hate actions. - Most content will be AI, and the net will in part die. Google search are already less than they were. - Everything will be subscription, and always connected will be required because there will be a need to check your subscription status. - EV's regulations will make it harder/impossible for people to buy cars, and fixing an ICE will also be super hard.


the always connected bit showed up during covid in Australia, because you had to have a smart phone ready to show people you scanned into everything.. including the grocery store.


Some parts of Australia.

I was here for most of the pandemic and don't even own or use a dumb mobile phone let alone a smart one. Wasn't an issue in the slightest.


At any point we are one catalyst event and few months away from civil and political rights as a subscription. Software professionals will be happy to implement the system for bonuses and public funding. Smartphone will be the enabler. Example - during covid the free movement within Schengen Area (look up how sophisticated and nuanced this arrangement is) was limited by vaccination certificate, which was... a PDF with qr code, or a screen within mobile app.


I actually checked my history then.

There weren't many real tech companies in Malaysia. Even FAANG was still unstable, nobody really expected them to last. Especially that Amazon dude who was never making any profits.

Most of the tech work here was just advanced spreadsheets. I thought the future was just boring stuff like better/faster/more scalable CRUD. There were a lot of poor jobs for intelligent people - basically if you graduated from Stanford, you'd probably end up selling foreign tech, building better spreadsheets, or consulting for a bank that doesn't care for tech.

What I would bet on then was optical routers and 5G. Much of the other stuff was just sci fi.

AI was one of those things. It's always been around as long as computers have; what are the odds someone would actually figure it out? It's more likely that some other paradigm like evolutionary programming would have taken off.


I expected augmented reality to be widespread and big market, like what AI is now. Didn't expect AI in any form. The most I could think of was OpenCV-based recognition and counting of traffic, and similar tasks.


Remembered that 3D printing still seemed the future of lots of stuff. But still 99% of houses aren't built that way.


I had literally zero predictions. I was too ignorant to have any.


Cheap Cloud.

I expected the cost of compute from major cloud providers to be a race to $0.

Never happened.

Compute from major cloud providers is still extremely expensive.

You only get those savings from ‘smaller’ cloud providers, like Hetner/OVH/etc - where it’s hard to even purchase a server, rack it and host it for the same cost they can. (Which is what I expected would happen with the majors)


- I didn't foresee crypto - I didn't foresee web3 per se but something similar - I somehow knew something like tik-tok or youtube shorts/IG reels would popup, e.g. "cheaper" video. I didn't think up of the "scroll down" UI.


How about 20 years ago? I saw Hinton's Restricted boltzmann machine and was dreaming of what we have today. Unsupervised learning, but in reality we got 10 years of supervised models before unsupervised was big.


In 2002 I was dreaming of having a voice-controlled smart teapot connected to water pipe, that would pour and boil me the exact amount I need (and other amount for others).

Now I dread this stuff and avoid buying devices that live their own life and make decisions for me (and are always buggy), plus require a whole lot of maintenance. (Maintenance and fragility are what comes to my mind when guys show their raspberry-pi home contraptions off.)


oh, and I forgot about digital signatures. In 2001 I made a presentation in university class about digital signatures, being impressed by PGP/GPG tech. Well, it is now used in some state services and chores like filing some request, or tax/statistics report, but people use the keys like just long passwords, and if you need anything serious -- come in person and use a pen.


The future is not evenly distributed. The only thing I have signed by hand in the last 4 years are package deliveries.

- Change of union, digital signature - New contract at work, digital signature - Moved apartment twice, digital signature - Signed my kid up for school, digital signature - New phone contract 5-6 times, digital signature - Made hundreds of payments online, digital signature. - updated my tax information, digital signature

All using NemID or later MitID here in Denmark.


Related, of you havent already seen these threads they are fun to read through: https://news.ycombinator.com/item?id=6994370


> Moore's Law remains halted. Nobody figures out how graphene will help. Programmers continue (see below) to move away from Java in favour of C to claw back some performance.

Hahaha. Electron probably makes them cry


Some stuff was pretty accurate. Most people got Apple wrong though.


I expected our current AI to be much further in the future, like 50 or 100 years.




Consider applying for YC's Spring batch! Applications are open till Feb 11.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: