5 ms·

This person is asking about career prospects 10-15 years out though. I'm sorry, but the landscape in then might be as alien to someone asking today, as todays

This person is asking about career prospects 10-15 years out though. I'm sorry, but the landscape in then might be as alien to someone asking today, as todays would have been to someone asking 15 years ago (2008). What John said is correct, but personally I think he's underplaying how much people could be affected. Those "product skills" take years of grinding to really sharpen, and in 15 years only a few people might actually be needed to apply them

csande17 4y ago
Was the tech landscape much different 10-15 years ago? This is a genuine question; the iPhone App Store was really the last "big thing" to happen to the industry in my mind, and it came out in 2008.
- raincole 4y ago
  No much different. If you were an intern Java programmer 10 years ago, it's totally possible that you're still a senior Java programmer today.
pharmakom 4y ago
2008 wasn’t so different was it? I guess the big new thing in work since then is the “gig economy”.
- onion2k 4y ago
  We've had LLMs for about 5 years so far in non-academic research. If we're talking 10 years out that means we're looking at tech that's about 1/3 through its development to date. Take any mature-ish technology that you use today and compare the version 1/3 through its life to the version you use now. Look at Chrome 20 compare to Chrome 111, or React 14 compared to React 18, or an iPhone 4 compared to an iPhone 14, or a car from 1950 compared to a car today... The difference is always quite significant. Superficially they're still the same thing, but if you look at the detail everything is just better. AI will be the same.
  - marginalia_nu 4y ago
    Eh, the difference is almost entirely in presentation. At the core a car today doesn't do significantly more than a car in 1962.
    - onion2k 4y ago
      Superficially they're similar in the "they both have 4 wheels and an engine" sense, but you could examine literally any part of a car today compared to one from the 1950s and find huge improvements. The efficiency, safety, comfort, tech, manufacturing... Everything is better.
      - microtonal 4y ago
        I agree with the safety angle, but besides that driving a car from 2023 is not substantially different from the 1950s (in the sense that it opens up a lot of new possibilities).
  - jstx1 4y ago
    You can't extrapolate from an arbitrary selection of technologies and assume that LLMs will have the same trajectory. They could be like the iPhone, or they could be like self-driving cars which have been a year away from replacing all drivers for 10 years now.
    - pharmakom 4y ago
      Self driving cars a few years ago seem particular close in hype level and apprehension to LLM today but progress on those has not matched expectations at all. What if GPT4 is the last major advance in LLMs for a really long time?
      - thequadehunter 4y ago
        Might just be me, but I think the big difference here is the level of adoption. Everybody with an internet connection can use an LLM. It hits closer to home that way, whereas driving is very dangerous and most people haven't used a self driving car before.
- pzo 4y ago
  The whole mobile economy pretty much started in 2008. First iPhone was released in 2007 but App Store was lunched in 2008. This changed landscape dramatically even if you consider software development. Before 2008 you were fine with writing just windows only desktop app in Delphi - no smartphones, tablets, smartwatches, smart tvs and could leave out supporting macOS or Linux
- pcthrowaway 4y ago
  Well I'm in web dev (though I was studying CS in 2008) and the 2008 landscape had almost none of the same things. jQuery was not yet a household name, let alone SPAs. Facebook had barely 100 million users. Marc Andreessen yet hadn't written about "software eating the world". Personally I was more optimistic. If anything, the last 15 years have seen the growth of an attitude of tech "entitlement" because hackers got to the a lot of the ideas that now seem obvious in hindsight before a lot of the big corps could. I'm sure there's still room for innovation, but I think a lot of it going forward will be driven by rapid improvement in AI capabilities. In 2008, tech wasn't everywhere. iphones were brand new and very few people had them.. There was no "mobile browser" market share (though we did have SMS gateways). 77% of the global population hadn't even been on the internet yet. AI looks like it's going to be at the forefront of the next big wave of fundamental changes to society, and it's really hard to predict where that will lead us. But I suspect it's going to become apparent that this relatively brief period of tech-elite empowerment was a historical anomaly, because the AI underlings are going to be willing to do a lot more work with none of the coddling, and they're going to improve very quickly.
  - revelio 4y ago
    Huh? Gmail was a spa and that dates from 2004. I don't think a VC writing a blog post says much about how tech had changed. Smartphones existed before the iPhone and Android, they just weren't as popular.
    - pcthrowaway 4y ago
      The term "SPA" wasn't in use until at least after 2009 and gmail was probably using some hacky AJAX (XMLHTTPRequest wasn't even standardized until, what, 2006?). Chrome wasn't launched until 2008 so they weren't able to get away with just adding the APIs they needed into the browser. Backbone wasn't even released until 2010 and Angular probably wasn't conceived of internally until late 2008. Yes, gmail might have had some SPA-like behaviour in 2004-2006 but it was nothing like what we have today. Pretty sure I got access in 2005 because I knew someone who worked at google, and it was mostly doing full refreshes between actions at the time, like pretty much the entire rest of the web
sneak 4y ago
You have plenty of time and can learn CS and earn a lot of money for years even if at exactly 120 months from now your job is made obsolete. It doesn't take 9 years to learn to code. The premise of all this seems to be that learning how to program computers is difficult or complex. It is not. Also, AI will not replace human reasoning in 10-15 years. If it does, it means AGI, and we all have much bigger problems than layoffs.
- Yoric 4y ago
  > It doesn't take 9 years to learn to code. True. But I guess the big question is what kind of skills you're going to need after that.
banyaaa 4y ago
I doubt AI will replace any job in my lifetime (got 40-50 years left). Progress will grind to a halt just like self driving cars did because the real world is just too chaotic and 'random' to be captured by a formula/equation/algorithm. My prediction is: AGI is theoretically possible, but would require impractical amounts of computing power - kinda like how intergalactic travel will never happen.
- pzo 4y ago
  We don’t AGI for LLM to be useful. And regrading comparison with self driving car they are still improving just the bar for them is much higher. If autopilot works 99.9% if time then 1 out of 1000 drivers will die - so technology has to be even better. for LLM is enough if it’s 90% good to be broadly useful.
  - AshamedCaptain 4y ago
    Not to "replace programmers", though. For the same reason you can't replace drivers even though they also make mistakes.
    - pzo 4y ago
      It’s not about replacing all programmers. If one programmer with AI assistant can do work the same as 2 programmers then one position is redundant. Even with self driving truck if one truck driver is leading another truck behind controlled by AI, and just for safety you have somewhere C&C center with one person monitoring 4 such ai trucks and in case unexpected event remotely take over control then one truck driver position is redundant.
      - ldhough 4y ago
        While I do think there is some threshold where increased productivity makes positions redundant, I don't think 2x would do it in most orgs. My current team easily has enough work for us all to be 2x more productive.
  - 4y ago
r_hoods_ghost 4y ago
2008 was extremely similar to today, although the webdev ecosystem wasn't quite as degenerate. I'd say you'd have to go back to the pre internet era to find a work environment that was fundamentally different.
chii 4y ago
> as todays would have been to someone asking 15 years ago (2008). i dont think, if you took someone from 15 yrs ago, and transplanted them here today, that they'd find it all that different technologically. Sure, machines are faster, slightly different, and such, but the fundamentals haven't changed. A software engineer could just as well write an app today as they had 15 yrs ago. You'd have to go back 30 yrs, for computers (and the landscape of computing) to have been different enough, that you can't transplant a software engineer.
- jocaal 4y ago
  The programmers back in the day were pretty good, i think a decent programmer from 30 years ago would be better that the average today.
- d0mine 4y ago
  30 years ago (1993): Linux existed, Python existed, web existed (mosaic), DOOM (3D graphics), and even Apple Newton (mobile) existed; and C, shell, windows (GUI), spreadsheet, sql, etc were known long before that. What exactly revolutionary happened in the last 30 years? javascript? (two weeks project) amazon, google, facebook, netflix, iphone, instagram, tiktok -- execution is great but seems inevitable that somebody will create it. Ok, for non-IT people iphone was a game changer (the first personal computer that your grandmother can actually use). The ability of generative AI to produce BS indistinguishable from human BS is very impressive but it remains to be seen whether it is a net positive for an average developer (the time wasted correcting it, waiting for its output can be spent understanding the problem better--the typing the code itself is a small part of a programmer who knows what they are doing).
dennis_jeeves1 4y ago
>What John said is correct, but personally I think he's underplaying how much people could be affected. Agreed, what John said was a bit of a platitude. I understand the spirit of what he said but he could have phrased it better.
scaramanga 4y ago
> I'm sorry, but the landscape in then might be as alien to someone asking today, as todays would have been to someone asking 15 years ago (2008). Hahahah. Yes. Who could have foreseen the trailblazing advances in the tech industry such as "television, but over the internet", "booking rooms, but via a website" or "posting messages on a forum" Don't forget the stuff powering it: "RPC, but over HTTP", "scripting languages, but compiled", or "Key-value stores" If only I had dared to dream.

csande17 4y ago

Was the tech landscape much different 10-15 years ago? This is a genuine question; the iPhone App Store was really the last "big thing" to happen to the industry in my mind, and it came out in 2008.

raincole 4y ago
No much different. If you were an intern Java programmer 10 years ago, it's totally possible that you're still a senior Java programmer today.

pharmakom 4y ago

2008 wasn’t so different was it? I guess the big new thing in work since then is the “gig economy”.

onion2k 4y ago
We've had LLMs for about 5 years so far in non-academic research. If we're talking 10 years out that means we're looking at tech that's about 1/3 through its development to date. Take any mature-ish technology that you use today and compare the version 1/3 through its life to the version you use now. Look at Chrome 20 compare to Chrome 111, or React 14 compared to React 18, or an iPhone 4 compared to an iPhone 14, or a car from 1950 compared to a car today... The difference is always quite significant. Superficially they're still the same thing, but if you look at the detail everything is just better. AI will be the same.
- marginalia_nu 4y ago
  Eh, the difference is almost entirely in presentation. At the core a car today doesn't do significantly more than a car in 1962.
  - onion2k 4y ago
    Superficially they're similar in the "they both have 4 wheels and an engine" sense, but you could examine literally any part of a car today compared to one from the 1950s and find huge improvements. The efficiency, safety, comfort, tech, manufacturing... Everything is better.
    - microtonal 4y ago
      I agree with the safety angle, but besides that driving a car from 2023 is not substantially different from the 1950s (in the sense that it opens up a lot of new possibilities).
- jstx1 4y ago
  You can't extrapolate from an arbitrary selection of technologies and assume that LLMs will have the same trajectory. They could be like the iPhone, or they could be like self-driving cars which have been a year away from replacing all drivers for 10 years now.
  - pharmakom 4y ago
    Self driving cars a few years ago seem particular close in hype level and apprehension to LLM today but progress on those has not matched expectations at all. What if GPT4 is the last major advance in LLMs for a really long time?
    - thequadehunter 4y ago
      Might just be me, but I think the big difference here is the level of adoption. Everybody with an internet connection can use an LLM. It hits closer to home that way, whereas driving is very dangerous and most people haven't used a self driving car before.
pzo 4y ago
The whole mobile economy pretty much started in 2008. First iPhone was released in 2007 but App Store was lunched in 2008. This changed landscape dramatically even if you consider software development. Before 2008 you were fine with writing just windows only desktop app in Delphi - no smartphones, tablets, smartwatches, smart tvs and could leave out supporting macOS or Linux
pcthrowaway 4y ago
Well I'm in web dev (though I was studying CS in 2008) and the 2008 landscape had almost none of the same things. jQuery was not yet a household name, let alone SPAs. Facebook had barely 100 million users. Marc Andreessen yet hadn't written about "software eating the world". Personally I was more optimistic. If anything, the last 15 years have seen the growth of an attitude of tech "entitlement" because hackers got to the a lot of the ideas that now seem obvious in hindsight before a lot of the big corps could. I'm sure there's still room for innovation, but I think a lot of it going forward will be driven by rapid improvement in AI capabilities. In 2008, tech wasn't everywhere. iphones were brand new and very few people had them.. There was no "mobile browser" market share (though we did have SMS gateways). 77% of the global population hadn't even been on the internet yet. AI looks like it's going to be at the forefront of the next big wave of fundamental changes to society, and it's really hard to predict where that will lead us. But I suspect it's going to become apparent that this relatively brief period of tech-elite empowerment was a historical anomaly, because the AI underlings are going to be willing to do a lot more work with none of the coddling, and they're going to improve very quickly.
- revelio 4y ago
  Huh? Gmail was a spa and that dates from 2004. I don't think a VC writing a blog post says much about how tech had changed. Smartphones existed before the iPhone and Android, they just weren't as popular.
  - pcthrowaway 4y ago
    The term "SPA" wasn't in use until at least after 2009 and gmail was probably using some hacky AJAX (XMLHTTPRequest wasn't even standardized until, what, 2006?). Chrome wasn't launched until 2008 so they weren't able to get away with just adding the APIs they needed into the browser. Backbone wasn't even released until 2010 and Angular probably wasn't conceived of internally until late 2008. Yes, gmail might have had some SPA-like behaviour in 2004-2006 but it was nothing like what we have today. Pretty sure I got access in 2005 because I knew someone who worked at google, and it was mostly doing full refreshes between actions at the time, like pretty much the entire rest of the web

sneak 4y ago

You have plenty of time and can learn CS and earn a lot of money for years even if at exactly 120 months from now your job is made obsolete. It doesn't take 9 years to learn to code. The premise of all this seems to be that learning how to program computers is difficult or complex. It is not. Also, AI will not replace human reasoning in 10-15 years. If it does, it means AGI, and we all have much bigger problems than layoffs.

Yoric 4y ago
> It doesn't take 9 years to learn to code. True. But I guess the big question is what kind of skills you're going to need after that.

banyaaa 4y ago

I doubt AI will replace any job in my lifetime (got 40-50 years left). Progress will grind to a halt just like self driving cars did because the real world is just too chaotic and 'random' to be captured by a formula/equation/algorithm. My prediction is: AGI is theoretically possible, but would require impractical amounts of computing power - kinda like how intergalactic travel will never happen.

pzo 4y ago
We don’t AGI for LLM to be useful. And regrading comparison with self driving car they are still improving just the bar for them is much higher. If autopilot works 99.9% if time then 1 out of 1000 drivers will die - so technology has to be even better. for LLM is enough if it’s 90% good to be broadly useful.
- AshamedCaptain 4y ago
  Not to "replace programmers", though. For the same reason you can't replace drivers even though they also make mistakes.
  - pzo 4y ago
    It’s not about replacing all programmers. If one programmer with AI assistant can do work the same as 2 programmers then one position is redundant. Even with self driving truck if one truck driver is leading another truck behind controlled by AI, and just for safety you have somewhere C&C center with one person monitoring 4 such ai trucks and in case unexpected event remotely take over control then one truck driver position is redundant.
    - ldhough 4y ago
      While I do think there is some threshold where increased productivity makes positions redundant, I don't think 2x would do it in most orgs. My current team easily has enough work for us all to be 2x more productive.
- 4y ago

r_hoods_ghost 4y ago

2008 was extremely similar to today, although the webdev ecosystem wasn't quite as degenerate. I'd say you'd have to go back to the pre internet era to find a work environment that was fundamentally different.

chii 4y ago

> as todays would have been to someone asking 15 years ago (2008). i dont think, if you took someone from 15 yrs ago, and transplanted them here today, that they'd find it all that different technologically. Sure, machines are faster, slightly different, and such, but the fundamentals haven't changed. A software engineer could just as well write an app today as they had 15 yrs ago. You'd have to go back 30 yrs, for computers (and the landscape of computing) to have been different enough, that you can't transplant a software engineer.

jocaal 4y ago
The programmers back in the day were pretty good, i think a decent programmer from 30 years ago would be better that the average today.
d0mine 4y ago
30 years ago (1993): Linux existed, Python existed, web existed (mosaic), DOOM (3D graphics), and even Apple Newton (mobile) existed; and C, shell, windows (GUI), spreadsheet, sql, etc were known long before that. What exactly revolutionary happened in the last 30 years? javascript? (two weeks project) amazon, google, facebook, netflix, iphone, instagram, tiktok -- execution is great but seems inevitable that somebody will create it. Ok, for non-IT people iphone was a game changer (the first personal computer that your grandmother can actually use). The ability of generative AI to produce BS indistinguishable from human BS is very impressive but it remains to be seen whether it is a net positive for an average developer (the time wasted correcting it, waiting for its output can be spent understanding the problem better--the typing the code itself is a small part of a programmer who knows what they are doing).

dennis_jeeves1 4y ago

>What John said is correct, but personally I think he's underplaying how much people could be affected. Agreed, what John said was a bit of a platitude. I understand the spirit of what he said but he could have phrased it better.

scaramanga 4y ago

> I'm sorry, but the landscape in then might be as alien to someone asking today, as todays would have been to someone asking 15 years ago (2008). Hahahah. Yes. Who could have foreseen the trailblazing advances in the tech industry such as "television, but over the internet", "booking rooms, but via a website" or "posting messages on a forum" Don't forget the stuff powering it: "RPC, but over HTTP", "scripting languages, but compiled", or "Key-value stores" If only I had dared to dream.