In fact, in my opinion, one of the benefits of AI tools that is often overlooked is "psychological support". When you are stuck at work, it will give you a push. Even if it is not completely right, it is enough to get you moving. The feeling of "no longer fighting alone at work" is actually more important than many people think.
pyman 46 minutes ago [-]
I asked my students to write a joke about AI. Sometimes humor is the best way to get people to talk about their fears without filters. One of them wrote:
"I went to work early that day and noticed my monitor was on, and code was being written without anyone pressing any keys. Something had logged into my machine and was writting code. I ran to my boss and told him my computer had been hacked. He looked at me, concerned, and said I was hallucinating. It's not a hacker, he said. It's our new agent. While you were sleeping, it built the app we needed. Remember that promotion you always wanted? Well, good news buddy! I'm promoting you to Prompt Manager. It's half the money, but you get to watch TikTok videos all day long!'"
Hard to find any real reassurance in that story.
psunavy03 15 hours ago [-]
"Great news, boss! We invented this new tool that allows nontechnical people to write code in English! Now anyone can deploy applications, and we don't have to hire all those expensive developers!"
"Wow, show it to me!"
"OK here it is. We call it COBOL."
musicale 4 hours ago [-]
FORTRAN (FORmula TRANslator) was another "AI" project in "automatic programming":
"Before 1954, almost all programming was done in machine language or assembly language. Programmers rightly regarded their work as a complex, creative art that required human inventiveness to produce an efficient program."
"The IBM Mathematical Formula Translating System or briefly, FORTRAN, will comprise a large set of programs to enable the IBM 704 to accept a concise formulation of a problem in terms of a mathematical notation and to produce automatically a high speed 704 program for the solution of the problem."
Fortran promised to eliminate debugging. In 2015, I taught React is a functional programming way to create very fast, bug free apps and the project manager found ways to push us to the hair-on-fire status quo.
But it still has been immensely useful and a durable paradigm, even though usage hasn't been exactly as thought.
iamflimflam1 2 hours ago [-]
I think what we forget is these high level languages did open up programming to people who would have been considered “nontechnical” back in the day.
ath92 4 minutes ago [-]
This is true, but:
- there are more programmers today than there were back then
- the best programmers are still those who would be considered technical back then
glimshe 15 hours ago [-]
You're joking but it's true. I'm sure you know that. SQL had similar claims... Declarative, say what you need and the computer will do for you. Also written in English.
ako 15 hours ago [-]
And compared to what we had before SQL, it is much easier to use, and a lot more people are able to use it.
noworriesnate 15 hours ago [-]
But software developers often struggle to use sql and prefer using ORMs or analytical APIs like polars; the people who excel at sql are typically not programmers, they’re data engineers, DBAs, analysts, etc.
Maybe a similar bifurcation will arise where there are vibe coders who use LLMs to write everything, and there are real engineers who avoid LLMs.
Maybe we’re seeing the beginning of that with the whole bifurcation of programmers into two camps: heavy AI users and AI skeptics.
ruszki 4 hours ago [-]
What you can achieve with the standard SQL is taught on universities. The whole package. I’ve never met a developer, who struggled with that. When you use ORMs you need to follow SQL’s logic anyway. People use ORMs to avoid painful data conversions. Not to avoid the logic. Data engineers, DBAs, analysts, etc excel in specific databases, not in “SQL”.
FranzFerdiNaN 2 hours ago [-]
Ive worked in BI and data engineering my whole career and I’ve met plenty of programmers who struggled immensely with SQL once it went further than select and group by. And don’t get me started about their database design skills. It’s way too often a disaster hidden behind “it works for the software so good enough”.
Im more surprised by software engineers who do know these things than by the ones who don’t.
maccard 1 hours ago [-]
I’ve worked with gameplay programmers who can’t do simple 3D math, c++ programmers who fundamentally don’t understand pointers, backend developers who didn’t understand globals were shared state and cause race conditions, etc.
It’s not that SQL is hard, it’s that for any discipline the vast majority of people don’t have a solid grasp of the tools they’re using. Ask most tradespeople about the underlying thing they’re working with and you’ll have the same problem.
adalacelove 4 hours ago [-]
I'm a developer and:
- I hate ORMs, they are the source for a lot of obscure errors behind layers and layers of abstractions.
- I prefer analytical APIs for technical reasons, not just the language.
Reasons:
- I can compose queries, which in turn makes them easier to decompose
- It's easier to spot errors
- I avoid parsing SQL strings
- It's easier to interact with the rest of the code, both functions and objects
If I need to make just a query I gladly write SQL
eru 2 hours ago [-]
Well, the problem in ORM is the O. Objection-orientation is just a worse way to organise your data and logic than relational algebra.
It's just a shame that many languages don't support relational algebra well.
We had relations as a datatype and all the relevant operations over them (like join) in a project I was working on. It was great! Very useful for expressing business logic.
tpm 1 hours ago [-]
The problem in ORM is the M, the mapping is always lossy and a leaky abstraction.
idiotsecant 5 hours ago [-]
'real' engineers can use SQL just fine. This is a strange position to take.
sanderjd 4 hours ago [-]
> But software developers often struggle to use sql
Is this true? It doesn't seem true to me.
nathanfig 15 hours ago [-]
Claude made this point while reviewing my blog for me: the mechanization of farms created a whole lot more specialization of roles. The person editing CAD diagrams of next year's combine harvester may not be a farmer strictly speaking, but farming is still where their livelihood comes from.
lipowitz 14 hours ago [-]
Removing jobs that could only be performed by those living near the particular fields with those that can be done anywhere makes jobs for the person willing to take the least satisfactory compensation for the most skill and work.
Working the summer fields was one of the least desirable jobs but still gave local students with no particular skills a good supplemental income appropriate for whichever region.
eru 2 hours ago [-]
Yes, but automating these away means that food becomes cheaper.
We increase the overall total prosperity with that automation.
dredmorbius 14 hours ago [-]
Strictly speaking, farming is where all our livelihoods come from, in the greatest part. We're all living off the surplus value of food production.
(Also of other food, energy, and materials sourcing: fishing, forestry, mining, etc.)
This was the insight of the French economist François Quesnay in his Tableau économique, foundation of the Physiocratic school of economics.
> Strictly speaking, farming is where all our livelihoods come from, in the greatest part. We're all living off the surplus value of food production.
I don't think farming is special here, because food isn't special. You could make exactly the same argument for water (or even air) instead of food, and all of a sudden all our livelihoods would derive ultimately from the local municipal waterworks.
Whether that's a reductio ad absurdum of the original argument, or a valuable new perspective on the local waterworks is left as an exercise to the reader.
14 hours ago [-]
ameliaquining 15 hours ago [-]
Is that really because of the English-esque syntax, rather than because it was a step forward in semantic expressivity? If SQL looked like, say, C#'s LINQ method syntax, would it really be harder to use?
9rx 15 hours ago [-]
> Is that really because of the English-esque syntax
Well, what we had before SQL[1] was QUEL, which is effectively the same as Alpha[2], except in "English". Given the previous assertion about what came before SQL, clearly not. I expect SQL garnered favour because it is tablational instead of relational, which is the quality that makes it easier to understand for those not heavy in the math.
[1] Originally known as SEQUEL, a fun word play on it claiming to be the QUEL successor.
[2] The godfather language created by Codd himself.
dmkolobov 2 hours ago [-]
Do you have any advice for understanding the difference between "relational" and "tablational"? I remember hearing something about how SQL is not really relational from my college professor, but we never really explored that statement.
AdieuToLogic 7 hours ago [-]
Before SQL became an industry standard, many programs which required a persistent store used things like ISAM[0], VISAM (a variant of ISAM[0]), or proprietary B-Tree libraries.
None of these had "semantic expressivity" as their strength.
> If SQL looked like, say, C#'s LINQ method syntax, would it really be harder to use?
Or QBE, "Query By Exemple", that was another try by IBM to make a query language directly usable by anyone.
kayodelycaon 9 hours ago [-]
SQL and many DSLs (JIRA…) are actually used by plenty of non-technical users. Anyone who wants to build their own reports and do basic data analysis has sufficient incentive to learn it.
They are very much the exception that proves the rule though.
veqq 14 hours ago [-]
Er, have you heard of datalog or Prolog? Declarative programming really does work. SQL was just... Botched.
sanderjd 4 hours ago [-]
I think SQL is better than datalog. I suspect this is one of those opinions that may be somewhat outside consensus on a forum like HN, but is strongly within the consensus more broadly.
glimshe 11 hours ago [-]
Yes. And I think SQL is actually pretty good for what it does. My point, as the parent's (I suppose) is that we've heard this "XYZ, which uses natural language, will kill software development" before.
dredmorbius 14 hours ago [-]
I'd long ago (1990s-era) heard that the original intent was that office secretaries would write their own SQL queries.
(I'd love for someone to substantiate or debunk this for me.)
bazoom42 14 hours ago [-]
Early on, programming was considered secretarial work.
AdieuToLogic 7 hours ago [-]
> Early on, programming was considered secretarial work.
Incorrect.
Encoding a program was considered secretarial work, not the act of programming itself. Over time, "encoding" was shortened to "coding."
This is why the industry term "coder" is a pejorative descriptor.
0points 1 hours ago [-]
> This is why the industry term "coder" is a pejorative descriptor.
It is not.
eru 2 hours ago [-]
> This is why the industry term "coder" is a pejorative descriptor.
For some people some of the time. I don't think that's true in general.
bitpush 15 hours ago [-]
Bravo. This is the exact sentiment I have, but you expressed in a way that I could never have.
Most people miss the fact that technical improvements increases the pie in a way that was not possible before.
When digital cameras became popular, everybody become a photographer. That only made the world better, and we got soo many more good photographers. Same with YouTube & creativity.
And same with coding & LLMs. World will have lots more of apps, and programmers.
munificent 14 hours ago [-]
> That only made the world better, and we got soo many more good photographers.
I disagree with the "only" part here. Imagine a distribution curve of photos with shitty photos on the left and masterpieces on the right and the height at the curve is how many photos there are to be seen at that quality.
The digital camera transition massively increased the height of the curve at all points. And thanks to things like better autofocus, better low light performance, and a radically faster iteration loop, it probably shift the low and middle ends to the right.
It even certainly increased the number number of breathtaking, life-changing photos out there. Digital cameras are game-changes for photographic journalists traveling in difficult locations.
However... the curve is so high now, the sheer volume of tolerably good photos so overwhelming, that I suspect that average person actually sees fewer great photos than they did twenty years ago. We all spend hours scrolling past nice-but-forgottable sunset shots on Instagram and miss out on the amazing stuff.
We are drowning in a sea of "pretty good". It is possible for there to be too much media. Ultimately, we all have a finite amount of attention to spend before we die.
barrenko 47 minutes ago [-]
My perspective on this meta - A) a lot of people seem to like taking pictures and it opened up a lot of new opportunities, but B) if you're a painter by nature, and maybe by profession, you may get hurt by this new photography thingy (financially hurt, which in primates equals hurt hurt). And liking painting does not entail liking photography and vice-versa. So C) It's just not the same thing. (C is the most opaque here and has the largest effect)
DavidPiper 9 hours ago [-]
Thank you for describing this so eloquently.
Meaning no disrespect to photographers, I'm starting to think that a probable outcome of all the AI investment is a sharp uptick in shovelware.
If we can get AIs to build "pretty good" things - or even just "pretty average" things - cheaply, then our app stores, news feeds, ad feeds, company directives, etc, will be continuously swamped with it.
eru 2 hours ago [-]
> Meaning no disrespect to photographers, I'm starting to think that a probable outcome of all the AI investment is a sharp uptick in shovelware.
You can use AI to filter out the shovelware, so you never have to see it.
shinedog 5 hours ago [-]
You hit this so hard it was impossible not to recognize. In every sense there is too much "ok" shit (in every media realm) that we cannot help but miss amazing stuff. Knowing that I don't have enough time for all the incredible things that technology has enabled crushes me.
test6554 7 hours ago [-]
Experts warn that at current production levels, the supply of dick pics may actually outpace demand in a couple decades.
DavidPiper 6 hours ago [-]
I was under the impression that supply already vastly outstrips demand.
eru 2 hours ago [-]
Demand is very unevenly distributed. I think they are appreciated on Grindr.
kjkjadksj 6 hours ago [-]
It affects even the competent photographer. How many times do you see that photographer with all the gear sit in front of a literal statue and fire off a 30 shot burst in 2 seconds? I don’t envy these pro photo editors either today in sports. I wonder how many shots they have to go through per touchdown from all the photographers at the end zone firing a burst until everyone stands up and throws the ball back at the ref? After a certain point you probably have to just close your eyes and pick one of the shots that looks almost identical to another 400. Not a job for analysis paralysis people. I guess it sure beats having to wait for the slide film to develop.
dotancohen 17 minutes ago [-]
The AI is already picking out the best photo in those 400-shot bursts.
And sometimes it is even combining elements from different photos: Alice had her eyes closed in this otherwise great shot, but in this other shot her eyes were open. A little touch-up and we've got the perfect photo.
socalgal2 3 hours ago [-]
don't you just let the AI pick? I'm only half joking. I thought that was a feature added to smartphones a year or two ago?
dijksterhuis 14 hours ago [-]
> That only made the world better
Did it?
people now stand around on dance floors taking photos and videos of themselves instead of getting on dancing and enjoying the music. to the point where clubs put stickers on phones to stop people from doing it.
people taking their phone out and videoing / photographing something awful happening, instead of doing something helpful.
people travel to remote areas where the population has been separated from humanity and do stupid things like leave a can of coke there, for view count.
it’s not made things better, it just made things different. whether that’s better or worse depends on your individual perspective for a given example.
so, i disagree. it hasn’t only made things better. it made some things easier. some things better. some things worse. some things harder.
someone always loses, something is always lost. would be good if more people in tech remembered that progress comes at a cost.
thangalin 13 hours ago [-]
> people now stand around on dance floors taking photos and videos of themselves instead of getting on dancing and enjoying the music. to the point where clubs put stickers on phones to stop people from doing it.
There are other types of dances where dancers are far more interested in the dance than selfies: Lindy Hop, Blues, Balboa, Tango, Waltz, Jive, Zouk, Contra, and West Coast Swing to name a few. Here are videos from the Blues dance I help organize where none of the dancers are filming themselves:
Thank you for sharing your social media videos as evidence in a rebuttal to "camera phones are not all good; they're ubiquitous use has negative implication too". So delicious...
VonTum 9 hours ago [-]
The irony!
Though, I'll grant that there's not really a way to argue this without showing videos
kjkjadksj 6 hours ago [-]
That sort of dancing is basically a sport. You have to learn it, you have to get good at it after you learned it, and it is cardio after all. I think op was talking more about what you see in the edm scene these days. Where basically people aren’t there to dance like the old days or sing along like other genres, they are there to see a certain DJ and then they will post clips from the entire set on their instagram story. And they can do this because the dancing they are doing at the edm show is super passive kind of dancing where you are just swaying a little so you can hold the phone stably at the same time. If you were dancing like how they’d dance at the edm concerts in the 90s all rolling on molly it would be like your blues swing where its just too physical to do anything but rave around flinging your arms all around shirtless and sweaty.
ttoinou 3 hours ago [-]
Look into contact impro and ecstatic dance : cellphones are forbidden and you can dance however you like it
skeeter2020 5 hours ago [-]
Live music sucks when you're trying to watch the show and some dumb-dumb is holding their phone above their head to shoot the entire show with low-light, bad angle & terrible sound. NO ONE is going to watch that, and you wrecked the experience for many people. Put your phone away and live in the present, please...
flashgordon 13 hours ago [-]
I would add one thing though. The pie definitely gets bigger - but i feel there is a period of "downsizing" that happens. I think this is becuase of lack of ideas. When you have tool that (say) 10xes your productivity, its not that bosses will have ideas to build 10x the number of things - they will just look to cut costs first (hello lack of imagination and high interest rates).
sarchertech 8 hours ago [-]
We’ve had many improvements that increased productivity at least as much as current LLMs, and I don’t think any of them ever temporarily caused downsizing in the total number of programmers.
pipes 14 hours ago [-]
I thought photographers don't get paid well anymore due market saturation and few skills required to get a good photo?
kjkjadksj 6 hours ago [-]
It is still as hard as its been to get a good photo. They had full auto film cameras that could take good photos in the 70s but the devil is always the edge cases and the subconscious ability to take an evenly exposed (in the Ansel Adams definition not auto camera exposure definition), well composed image at the decisive moment. Understanding how lighting works (either natural, or different artificial light like flash or studio lighting) is also not easy.
It is pretty hard to break out but people still make names for themselves either from experience on assignments like the old days but also from instagram and other social media followings. People still need weddings shot and professional portraits taken which takes some skill in understanding the logistics of how to actually do that job well efficiently and managing your equipment.
bluefirebrand 10 hours ago [-]
> World will have lots more of apps, and programmers.
This is actually bad for existing programmers though?
Do you not see how this devalues your skills?
platevoltage 10 hours ago [-]
I see your point, but I'm having personally having a different experience.
A client of mine has gotten quite good at using Bolt and Lovable. He has since put me on 3 more projects that he dreamed up and vibe coded that would just be a figment of his imagination pre-AI.
He knows what's involved in software development, and knows that he can't take it all the way with these tools.
sarchertech 8 hours ago [-]
There are far more programmers now than in 1980, yet the average programmer makes far more (inflation adjusted) now.
kjkjadksj 6 hours ago [-]
Thank the Bangalore office for that.
FirmwareBurner 3 hours ago [-]
How much online shopping could you do from your PC in 1980? How many people had smartphones in 1980?
That's why sw devs salaries went up like crazy in our time and not in 1980.
But what new tech will we have, that will push the SW dev market demand up like internet connected PCs and smartphones did? All I see is stagnation in the near future, just maintaining or rewriting the existing shit that we have, not expanding into new markets.
bitpush 10 hours ago [-]
In the current state, yes. But that is also an opportunity, isn't it?
When online flight bookings came about, travel agents were displaced. The solution isn't "let's stop online flight bookings sites and protect travel agents" because that's an inefficient system
dijksterhuis 8 hours ago [-]
Why does every system need to be efficient?
hackernoops 7 hours ago [-]
Fractional reserve lending, rehypothecation, etc.
komali2 6 hours ago [-]
Under capitalism, because greater margins. Under not-capitalism, so as to free up resources and labor for other things or just increase available downtime for people.
lupire 7 hours ago [-]
Sorry to be that guy, but would to prefer if your computer and phone each cost $5000?
tonyedgecombe 3 hours ago [-]
In some ways I would, computing lost something once normal people were allowed in.
20after4 15 hours ago [-]
And now the business of wedding / portrait photographer has become hyper-competitive. Now everyone's cousin is an amateur photographer and every phone has an almost acceptable camera built in. It is much more difficult to have a profitable photography business compared to 20 years ago.
bachmeier 15 hours ago [-]
That's good to hear. Back when I got married there were some real jerks in the wedding photography business, and they weren't worried about running out of customers. Here's an actual conversation I had with one of them:
Me: "I'm getting married on [date] and I'm looking for a photographer."
Them, in the voice of Nick Burns: "We're already filling up for next year. Good luck finding a photographer this year."
Me: "I just got engaged. You never have anything open up?"
Them: "No" and hang up the phone.
The faster guys like that struggle to make a living, the better.
tonyedgecombe 3 hours ago [-]
I know a couple of professional photographers and neither of them will do weddings. It seems many of the clients are as bad as the photographers.
LargeWu 12 hours ago [-]
In the same breath, those photographers will complain about all the "amateurs" devaluing their services.
NewsaHackO 14 hours ago [-]
Definitely. What matters more is that the ability to take photos is available to more people, which is a net positive.
insane_dreamer 6 hours ago [-]
> everybody become a photographer. That only made the world better, and we got soo many more good photographers.
Not sure I agree. I haven't seen much evidence of "better photography" now that it's digital instead of film. There are a million more photos taken, yes, because the cost is zero. But quantity != quality or "better", and if you're an average person, 90% those photos are in some cloud storage and rarely looked at again.
You could argue that drones have made photography better because it's enabled shots that were impossible or extremely difficult before (like certain wildlife/nature shots).
One thing digital photography did do is decimate the photographer profession because there is so much abundance of "good enough" photos - why pay someone to take good ones? (This may be a lesson for software development too.)
deanCommie 2 hours ago [-]
> That only made the world better, and we got soo many more good photographers. Same with YouTube & creativity.
I think you really missed the point of what these technologies and innovations actually did for society and how it applies to today, underneath the snark.
In the 1970's, if you got gifted a camera, and were willing to put in the work to figure out how to use it, you learned a skill that immediately put you in rare company.
With enough practice of that skill you could be a professional photographer, which would be a good , reliable, well paid job. Now, the barrier of entry is nothing, so it's extremely competitive to be a professional photographer, and even the ones that succeed just scrape by. And you have to stand out on other things than the technical ability to operate a camera.
That's...what's about to happen (if it hasn't already) with software developers.
platevoltage 10 hours ago [-]
Fast forward a couple decades and "Ok here it is. We call it Dreamweaver"
and the statement is still true, except that many devs nowadays are JS-only, and are too scared or lazy as shit to learn another, relatively simple language like SQL. ("it's too much work". wtf do you think a job is. it's another name for work.)
because, you know, "we have to ship yesterday" (which funnily enough, is always true, like "tomorrow never comes").
8note 5 hours ago [-]
SQL is straightforward enough, but its not the sketchy part. taking down the database so other people cant use it by running a test query is the bad part.
the explains are not nearly as straightforward to read, and the process of writing SQL is to write the explain yourself, and then try to coax the database into turning SQL you write into that explain. its a much less pleasent LLM chat experience
ashoeafoot 2 hours ago [-]
We now have assembler, now anyone can program.
No, wait it was called natural language coding, now anyone can code.
No, wait it was called run anything self fixing code.
No wait, simplified domain specific language.
No, wait it was uml based coding.
No, wait excel makros.
No, wait its node based drag and drop .
No, wait its LLMs.
The mental retardation of no code is strong with the deciding caste, every reincarnation must be taxed.
chii 2 hours ago [-]
The big difference with LLM is that you don't have to have a conherant and logical thought, and the LLM will "fix" that for you by morphing it into the nearest coherent expression and show you the result.
Presumably, the LLM user will have sufficient brain capacity to verify that the result works as they have imagined (however incomplete the mental picture might be). They then have an opportunity to tweak, in real time (of sorts), to make the output closer to what they want. Repeat this as many times as needed/time available, and the output gets to be quite sufficient for purpose.
This is how traditional, bespoke software development would've worked with contractor developers. Except with LLM, the turnaround time is in minutes, rather than in days or weeks.
soulofmischief 2 hours ago [-]
What's wrong with visual programming?
michaelteter 9 hours ago [-]
Having experienced several overhyped corporate knee-jerk (and further press-amplified) silver bullets, I expect this will play out about as well as the previous ones.
And by that, I mean corps will make poor decisions that will be negative for thought workers while never really threatening executive compensation.
I see this latest one somewhat like TFA author: this is a HUGE opportunity for intelligent, motivated builders. If our jobs are at risk now or have already been lost, then we might as well take this time to make some of the things we have thought about making before but were too busy to do (or too fatigued).
In the process, we may not only develop nice incomes that are independent of PHB decisions, but some will even build things that these same companies will later want to buy for $$$.
dotancohen 18 seconds ago [-]
I've already started.
I've been recording to myself voice notes for years. Until now they've seemingly been near-read-only. The friction for recording them is often low (in settings where I can speak freely) but getting the information out of them has been difficult.
I'm now writing software to help me quickly get information out of the voice notes. So they'll be useful to me too, not just to future historians who happen upon my hard drive. I would not be able to devote the time to this without AI, even though most of the code and all the architecture is my own.
eqvinox 32 minutes ago [-]
Metaphors are fun, they "feel" meaningful, but… you still need to back that up.
> mechanized farm equipment
Sure, that could be a valid analogy.
Or maybe we invented CAD software for mechanical engineering, where we were making engineering drawings by hand before?
And that doesn't quite ring the same way in terms of obsoleting engineers…
nathanfig 18 hours ago [-]
Hi all - I write a lot for myself but typically don't share, hence the stream-of-consciousness style.
But I thought this might be worth blogifying just for the sake of adding some counter-narrative to the doomerism I see a lot regarding the value of software developers. Feel free to tear it apart :)
tasuki 1 hours ago [-]
I clicked only because I disagreed with the title. What a joy of an essay!
randfish 17 hours ago [-]
Thought it was great. Thanks for writing and submitting!
nathanfig 16 hours ago [-]
Thanks!
layer8 15 hours ago [-]
The humor was refreshing. :)
NoPicklez 6 hours ago [-]
My take just purely based on the title, I'm in the security space not a developer but I did study it during my degree.
I would say that when the fundamentals are easier to learn it becomes a great time to learn anything. I remember spending so much of my degree during software development trying to fix bugs and have things explained by trawling through online forums like many of us have. Looking for different ways of having concepts explained to me and how to apply them.
LLM's give us a fairly powerful tool to act as a sort of tutor in asking questions, feedback on code blocks, understanding concepts, where my code went wrong etc. Asking it all of the dumb questions we go trawling for.
But I can't speak to how this translates when you're a more intermediate developer.
el_benhameen 4 hours ago [-]
I have found them quite helpful in the same way. I can bounce ideas off of them or say “here’s my understanding of this; in what ways am I incorrect?”. I don’t trust them to have pinpoint accuracy on complex problems, but given the way they’re trained, I do trust them to be directionally correct. That makes getting past hangups faster and leads me to ask more, better questions of myself, which I think means I learn faster.
boxed 57 minutes ago [-]
> And like mechanized farm equipment, LLMs are cheap, plentiful, getting smaller every day, and- most importantly- require no training to operate.
I... assume that was meant sarcastically, but it's not at all clear from context I think.
Leo-thorne 1 hours ago [-]
Now really feels like a good time to start learning how to code. I used to get completely lost reading documentation, but with Copilot, I just type a few lines and it helps fill in the logic. It feels like having a more experienced person sitting next to me.
That said, I still try to figure out the logic myself first, then let AI help polish or improve it. It is a bit slower, but when something breaks, at least I know why.
AI has definitely lowered the barrier. But whether you can actually walk through the door still depends on you.
dehrmann 15 hours ago [-]
The farming quote is interesting, but one of the Jevons paradox requirements is a highly elastic demand curve, and food is inelastic.
The open questions right now are how much of a demand is there for more software, and where do AI capabilities plateau.
9rx 14 hours ago [-]
Either way, as quite visibility seen by all the late-1800s mansions still lining the country roads, the era of farmers being "overpaid", as the link puts it, came about 50-75 years after the combine was invented. If the metaphor is to hold, we can assume that developers are currently poor as compared to what the LLM future holds for them.
But, there is a key distinction that we would be remiss to not take note of: By definition, farmers are the owners of the business. Most software developers aren't owners, just lowly employees. If history is to repeat, it is likely that, as usual, the owners are those who will prosper from the advancement.
2 hours ago [-]
slt2021 14 hours ago [-]
demand for food is very elastic. if beef becomes more expensive, cheaper protein options get more demand (chicken, pork, tofu, beans).
fruits and all non-essential food items are famously very elastic, and constitute large share of the spending.
for example: if cheap cereal becomes abundant, it is only at the cost of poor quality, so demand for high quality cereal will increase.
the LLM driven software engineering will continuously increase the bar for quality and demand for high quality software
fulafel 2 hours ago [-]
Demand for eaten calories is not very elastic but plentiful food crops lead to piping crops through the wasteful, environmentally harmful & unethical thingthat is meat production.
giraffe_lady 15 hours ago [-]
Reported numbers vary but household food waste seems to be fairly high in developed economies, so food demand might be more elastic than intuition would expect.
dredmorbius 14 hours ago [-]
I've seen consistent values for food waste reported for at least the past 40 years, if not the past 80, in various sources. I suspect it's something of a constant. One observation I've seen is that food wastage now occurs far later in the processing cycle, which is to say, after far more resources (transport, processing, refrigeration, cooking) have been invested in it.
In the long term, food demand is elastic in that populations tend to grow.
kwk1 14 hours ago [-]
Perhaps we should say something like "food demand has an elasticity floor."
giraffe_lady 12 hours ago [-]
For sure.
ByteDrifter 54 minutes ago [-]
Reading this reminded me how much the learning curve is flattening. You can now learn by doing and debugging AI output. That’s a very different entry point from five years ago. Less lonely, more interactive.
rossdavidh 9 hours ago [-]
All of this is good reason that orgs _shouldn't_ be laying off developers, but none of it is a reason that they won't/aren't. In any case, I see more "if they're remote why can't they be on the low-wage side of the planet" at the moment, than I do "use AI instead of a developer", although they are no doubt related.
The more awkward truth is that most of what developers have been paid to do in the 21st century was, from the larger perspective, wasted. We mostly spent a lot of developer time in harvesting attention, not in actually making anything truly useful.
MichaelZuo 9 hours ago [-]
How does that follow…?
Most organizations do derive net benefit from laying off the below average and hiring the above average for a given compensation range, as long as the turnover is not too high.
And this delta increases when the above average can augment themselves more effectively, so it seems we should expect an even more intense sorting.
abalashov 15 hours ago [-]
I'm not sure if I agree with every aspect of the framing here; specifically, I don't think the efficiency gains are anywhere on par with a combine harvester.
However, I do agree that the premium shifts from mere "coding" ability -- we already had a big look into this with the offshoring wave two decades ago -- to domain expertise, comprehension of the business logic, ability to translate fluidly between different kinds of technical and nontechnical stakeholders, and original problem-solving ability.
nathanfig 15 hours ago [-]
Yeah I think the combine-harvester analogy is tempting because it's so easy to visualize how wheat can scale over a big square field and project that visual onto lines of code generated on a big square screen... forgetting that lines-of-code-generated is not inherently useful.
temporallobe 13 hours ago [-]
Essentially it’s the same as it always was. Back in the day, Low-code or No-code solutions implemented by non-technical people have always resulted in engineers having to come in behind them to clean up their mess. I’ve had quite the lucrative career doing just that.
6 hours ago [-]
nathanfig 13 hours ago [-]
Yeah, with current-state AI I foresee more such opportunities.
Ekaros 12 hours ago [-]
I think I will have good while in security. That is pointing all the mistakes and faults... And telling why something AI came up might not fully solve the problem.
So much room left. As I doubt every developer will double check things every time by asking.
6 hours ago [-]
prisenco 14 hours ago [-]
Upwork is already filling up with people who have vibe-coded their way into a pit and need experienced developers to pull them out.
billy99k 14 hours ago [-]
You can find good contract on Upwork, but you need to go through lots of bad ones. I find around 5 good contracts there per year. I find that even when a client agrees on a rate, Upwork has the reputation of finding inexpensive workers, and you will get many clients trying to pay you less.
I'm also a bit tired of running into people that are 'starting a contracting firm' and have 0 clients or direction yet and just want to waste your time.
nathanfig 13 hours ago [-]
Really! That could make for some really interesting stories. Fascinating to think of LLMs as a customer acquisition pipeline for developers.
platevoltage 10 hours ago [-]
I've snagged at least one of them.
karczex 17 hours ago [-]
It's like "we invented Fortran so there will be no need for so many developers"
nathanfig 16 hours ago [-]
An interesting parallel because there were undoubtedly some people who worried we would lose something important in the craft of instruction-level programming, and almost certainly we have in relative terms. But in absolute numbers I am confident we have more low-level programmers than we did before Fortran.
And if I were to jump into instruction-level programming today I would start by asking an LLM where to begin...
marcosdumay 14 hours ago [-]
Fortran was a much larger jump in productivity than agentic coding...
yodsanklai 15 hours ago [-]
> What do you do while awaiting the agents writing your code?
I browse the web. Eventually, I review the agent code and more often than not, I rewrite it.
I also remember this! Maybe a subconscious influence
irrational 4 hours ago [-]
> and historical romance novels will rightly remember us as rugged and sexy.
Damn straight we are.
agentultra 5 hours ago [-]
If you’re going to use LLMs to learn software development, great! Welcome!
Just, don’t skip out on learning the fundamentals. There’s no royal road to knowledge and skill. No shortcuts. No speed running, downloading kung fu, no passing go.
Why?
Because the only thing LLMs do is hallucinate. Often what they generate is what you’re looking for. It’s the right answer!
But if you don’t know what and L1 cache is or how to lay out data for SIMD; no amount of yelling at the bot is going to fix the poor performance, the security errors, and the logic errors. If you don’t know what to ask you won’t know what you’re looking at. And you won’t know how to fix it.
So just remember to learn the fundamentals while you’re out there herding the combine space harvesters… or whatever it is kids do these days.
everyone 2 hours ago [-]
I learned programming in 2013, I was asking questions on stack overflow constantly while learning, people there were super friendly and supportive and answered my questions. SO was pivotal for me learning so fast.. Ive been a game programmer ever since.. This year I learned web-dev and made my 1st commercial web app. SO is totally dead though, utterly useless these days, but I used chatGPT to fill the same role and it worked great. Its a shame about SO though.
SeanDav 16 hours ago [-]
>> "ChadGPT"
There actually is a ChadGPT but I assume the OP meant ChatGPT
nathanfig 16 hours ago [-]
Oh I should have known - yeah I was just being facetious
rr808 7 hours ago [-]
The management at my corporate job literally say in our townhalls that they expect AI to increase productivity and reduce costs. Makes logical sense to me, the glory days of high wages are over.
vincenthwt 6 hours ago [-]
Are you talking about the high wages of software engineers or management? Makes sense to me— the glory days of high management and CEO salaries are over.
revskill 3 hours ago [-]
The hardest part is debugging.
ed_mercer 54 minutes ago [-]
In theory one should be able to hand this over to a MCP debugging server.
16 hours ago [-]
5 hours ago [-]
freekh 16 hours ago [-]
Nice article! Reflects my views as well!
alganet 14 hours ago [-]
> and now with far greater reach and speed than ever before
I heard that before. Borland Delphi, Microsoft FrontPage, Macromedia Flash and so on. I learned how in 5 years or so, these new technologies would dominate everything.
Then I learned that two scenarios exist. One of them is "being replaced by a tool", the other is "being orphaned by a tool". You need to be prepared for both.
nathanfig 13 hours ago [-]
Yes, if you built your career on FrontPage you have probably had a bad time. Many such cases.
That said, even if the specific products like Cursor or ChatGPT are not here in 5 years, I am confident we are not going to collectively dismiss the utility of LLMs.
alganet 12 hours ago [-]
I can see it being useful for summarization, or creative writing. What makes you so sure that LLMs will be useful _for programming_ in the long run?
fuzztester 8 hours ago [-]
[flagged]
mirkodrummer 10 hours ago [-]
> LLMs really are like combine harvesters; allowing one to do the work of many.
Heck I'm so tired of statements like this, many who? It's already a lot an LLM that automate/help the boring/tedious part of my job, I have yet to see taking over 2, 5 or 10 of my collegues, just knowing what a hawful lot these tiredlessly dudes do I couldn't ever imagine doing also their job. imo such statements have very short shelf life
"I went to work early that day and noticed my monitor was on, and code was being written without anyone pressing any keys. Something had logged into my machine and was writting code. I ran to my boss and told him my computer had been hacked. He looked at me, concerned, and said I was hallucinating. It's not a hacker, he said. It's our new agent. While you were sleeping, it built the app we needed. Remember that promotion you always wanted? Well, good news buddy! I'm promoting you to Prompt Manager. It's half the money, but you get to watch TikTok videos all day long!'"
Hard to find any real reassurance in that story.
"Wow, show it to me!"
"OK here it is. We call it COBOL."
"Before 1954, almost all programming was done in machine language or assembly language. Programmers rightly regarded their work as a complex, creative art that required human inventiveness to produce an efficient program."
-John Backus, "The History of Fortran I, II, and III", https://dl.acm.org/doi/10.1145/800025.1198345
"The IBM Mathematical Formula Translating System or briefly, FORTRAN, will comprise a large set of programs to enable the IBM 704 to accept a concise formulation of a problem in terms of a mathematical notation and to produce automatically a high speed 704 program for the solution of the problem."
-IBM, "Specifications for the IBM Mathematical FORmula TRANslating System, FORTRAN", http://archive.computerhistory.org/resources/text/Fortran/10...
"FORTRAN should virtually eliminate coding and debugging" https://news.ycombinator.com/item?id=3970011
But it still has been immensely useful and a durable paradigm, even though usage hasn't been exactly as thought.
Maybe a similar bifurcation will arise where there are vibe coders who use LLMs to write everything, and there are real engineers who avoid LLMs.
Maybe we’re seeing the beginning of that with the whole bifurcation of programmers into two camps: heavy AI users and AI skeptics.
Im more surprised by software engineers who do know these things than by the ones who don’t.
It’s not that SQL is hard, it’s that for any discipline the vast majority of people don’t have a solid grasp of the tools they’re using. Ask most tradespeople about the underlying thing they’re working with and you’ll have the same problem.
Reasons: - I can compose queries, which in turn makes them easier to decompose - It's easier to spot errors - I avoid parsing SQL strings - It's easier to interact with the rest of the code, both functions and objects
If I need to make just a query I gladly write SQL
It's just a shame that many languages don't support relational algebra well.
We had relations as a datatype and all the relevant operations over them (like join) in a project I was working on. It was great! Very useful for expressing business logic.
Is this true? It doesn't seem true to me.
Working the summer fields was one of the least desirable jobs but still gave local students with no particular skills a good supplemental income appropriate for whichever region.
We increase the overall total prosperity with that automation.
(Also of other food, energy, and materials sourcing: fishing, forestry, mining, etc.)
This was the insight of the French economist François Quesnay in his Tableau économique, foundation of the Physiocratic school of economics.
> Strictly speaking, farming is where all our livelihoods come from, in the greatest part. We're all living off the surplus value of food production.
I don't think farming is special here, because food isn't special. You could make exactly the same argument for water (or even air) instead of food, and all of a sudden all our livelihoods would derive ultimately from the local municipal waterworks.
Whether that's a reductio ad absurdum of the original argument, or a valuable new perspective on the local waterworks is left as an exercise to the reader.
Well, what we had before SQL[1] was QUEL, which is effectively the same as Alpha[2], except in "English". Given the previous assertion about what came before SQL, clearly not. I expect SQL garnered favour because it is tablational instead of relational, which is the quality that makes it easier to understand for those not heavy in the math.
[1] Originally known as SEQUEL, a fun word play on it claiming to be the QUEL successor.
[2] The godfather language created by Codd himself.
None of these had "semantic expressivity" as their strength.
> If SQL looked like, say, C#'s LINQ method syntax, would it really be harder to use?
Yes.
0 - https://en.wikipedia.org/wiki/ISAM
They are very much the exception that proves the rule though.
(I'd love for someone to substantiate or debunk this for me.)
Incorrect.
Encoding a program was considered secretarial work, not the act of programming itself. Over time, "encoding" was shortened to "coding."
This is why the industry term "coder" is a pejorative descriptor.
It is not.
For some people some of the time. I don't think that's true in general.
Most people miss the fact that technical improvements increases the pie in a way that was not possible before.
When digital cameras became popular, everybody become a photographer. That only made the world better, and we got soo many more good photographers. Same with YouTube & creativity.
And same with coding & LLMs. World will have lots more of apps, and programmers.
I disagree with the "only" part here. Imagine a distribution curve of photos with shitty photos on the left and masterpieces on the right and the height at the curve is how many photos there are to be seen at that quality.
The digital camera transition massively increased the height of the curve at all points. And thanks to things like better autofocus, better low light performance, and a radically faster iteration loop, it probably shift the low and middle ends to the right.
It even certainly increased the number number of breathtaking, life-changing photos out there. Digital cameras are game-changes for photographic journalists traveling in difficult locations.
However... the curve is so high now, the sheer volume of tolerably good photos so overwhelming, that I suspect that average person actually sees fewer great photos than they did twenty years ago. We all spend hours scrolling past nice-but-forgottable sunset shots on Instagram and miss out on the amazing stuff.
We are drowning in a sea of "pretty good". It is possible for there to be too much media. Ultimately, we all have a finite amount of attention to spend before we die.
Meaning no disrespect to photographers, I'm starting to think that a probable outcome of all the AI investment is a sharp uptick in shovelware.
If we can get AIs to build "pretty good" things - or even just "pretty average" things - cheaply, then our app stores, news feeds, ad feeds, company directives, etc, will be continuously swamped with it.
You can use AI to filter out the shovelware, so you never have to see it.
And sometimes it is even combining elements from different photos: Alice had her eyes closed in this otherwise great shot, but in this other shot her eyes were open. A little touch-up and we've got the perfect photo.
Did it?
people now stand around on dance floors taking photos and videos of themselves instead of getting on dancing and enjoying the music. to the point where clubs put stickers on phones to stop people from doing it.
people taking their phone out and videoing / photographing something awful happening, instead of doing something helpful.
people travel to remote areas where the population has been separated from humanity and do stupid things like leave a can of coke there, for view count.
it’s not made things better, it just made things different. whether that’s better or worse depends on your individual perspective for a given example.
so, i disagree. it hasn’t only made things better. it made some things easier. some things better. some things worse. some things harder.
someone always loses, something is always lost. would be good if more people in tech remembered that progress comes at a cost.
There are other types of dances where dancers are far more interested in the dance than selfies: Lindy Hop, Blues, Balboa, Tango, Waltz, Jive, Zouk, Contra, and West Coast Swing to name a few. Here are videos from the Blues dance I help organize where none of the dancers are filming themselves:
* https://www.facebook.com/61558260095218/videos/7409340551418...
* https://www.facebook.com/reel/3659488930863692
Though, I'll grant that there's not really a way to argue this without showing videos
It is pretty hard to break out but people still make names for themselves either from experience on assignments like the old days but also from instagram and other social media followings. People still need weddings shot and professional portraits taken which takes some skill in understanding the logistics of how to actually do that job well efficiently and managing your equipment.
This is actually bad for existing programmers though?
Do you not see how this devalues your skills?
A client of mine has gotten quite good at using Bolt and Lovable. He has since put me on 3 more projects that he dreamed up and vibe coded that would just be a figment of his imagination pre-AI.
He knows what's involved in software development, and knows that he can't take it all the way with these tools.
That's why sw devs salaries went up like crazy in our time and not in 1980.
But what new tech will we have, that will push the SW dev market demand up like internet connected PCs and smartphones did? All I see is stagnation in the near future, just maintaining or rewriting the existing shit that we have, not expanding into new markets.
When online flight bookings came about, travel agents were displaced. The solution isn't "let's stop online flight bookings sites and protect travel agents" because that's an inefficient system
Me: "I'm getting married on [date] and I'm looking for a photographer."
Them, in the voice of Nick Burns: "We're already filling up for next year. Good luck finding a photographer this year."
Me: "I just got engaged. You never have anything open up?"
Them: "No" and hang up the phone.
The faster guys like that struggle to make a living, the better.
Not sure I agree. I haven't seen much evidence of "better photography" now that it's digital instead of film. There are a million more photos taken, yes, because the cost is zero. But quantity != quality or "better", and if you're an average person, 90% those photos are in some cloud storage and rarely looked at again.
You could argue that drones have made photography better because it's enabled shots that were impossible or extremely difficult before (like certain wildlife/nature shots).
One thing digital photography did do is decimate the photographer profession because there is so much abundance of "good enough" photos - why pay someone to take good ones? (This may be a lesson for software development too.)
I think you really missed the point of what these technologies and innovations actually did for society and how it applies to today, underneath the snark.
In the 1970's, if you got gifted a camera, and were willing to put in the work to figure out how to use it, you learned a skill that immediately put you in rare company.
With enough practice of that skill you could be a professional photographer, which would be a good , reliable, well paid job. Now, the barrier of entry is nothing, so it's extremely competitive to be a professional photographer, and even the ones that succeed just scrape by. And you have to stand out on other things than the technical ability to operate a camera.
That's...what's about to happen (if it hasn't already) with software developers.
today:
s/COBOL/SQL
and the statement is still true, except that many devs nowadays are JS-only, and are too scared or lazy as shit to learn another, relatively simple language like SQL. ("it's too much work". wtf do you think a job is. it's another name for work.)
because, you know, "we have to ship yesterday" (which funnily enough, is always true, like "tomorrow never comes").
the explains are not nearly as straightforward to read, and the process of writing SQL is to write the explain yourself, and then try to coax the database into turning SQL you write into that explain. its a much less pleasent LLM chat experience
No, wait it was called natural language coding, now anyone can code.
No, wait it was called run anything self fixing code. No wait, simplified domain specific language.
No, wait it was uml based coding.
No, wait excel makros.
No, wait its node based drag and drop .
No, wait its LLMs.
The mental retardation of no code is strong with the deciding caste, every reincarnation must be taxed.
Presumably, the LLM user will have sufficient brain capacity to verify that the result works as they have imagined (however incomplete the mental picture might be). They then have an opportunity to tweak, in real time (of sorts), to make the output closer to what they want. Repeat this as many times as needed/time available, and the output gets to be quite sufficient for purpose.
This is how traditional, bespoke software development would've worked with contractor developers. Except with LLM, the turnaround time is in minutes, rather than in days or weeks.
And by that, I mean corps will make poor decisions that will be negative for thought workers while never really threatening executive compensation.
I see this latest one somewhat like TFA author: this is a HUGE opportunity for intelligent, motivated builders. If our jobs are at risk now or have already been lost, then we might as well take this time to make some of the things we have thought about making before but were too busy to do (or too fatigued).
In the process, we may not only develop nice incomes that are independent of PHB decisions, but some will even build things that these same companies will later want to buy for $$$.
I've been recording to myself voice notes for years. Until now they've seemingly been near-read-only. The friction for recording them is often low (in settings where I can speak freely) but getting the information out of them has been difficult.
I'm now writing software to help me quickly get information out of the voice notes. So they'll be useful to me too, not just to future historians who happen upon my hard drive. I would not be able to devote the time to this without AI, even though most of the code and all the architecture is my own.
> mechanized farm equipment
Sure, that could be a valid analogy.
Or maybe we invented CAD software for mechanical engineering, where we were making engineering drawings by hand before?
And that doesn't quite ring the same way in terms of obsoleting engineers…
But I thought this might be worth blogifying just for the sake of adding some counter-narrative to the doomerism I see a lot regarding the value of software developers. Feel free to tear it apart :)
I would say that when the fundamentals are easier to learn it becomes a great time to learn anything. I remember spending so much of my degree during software development trying to fix bugs and have things explained by trawling through online forums like many of us have. Looking for different ways of having concepts explained to me and how to apply them.
LLM's give us a fairly powerful tool to act as a sort of tutor in asking questions, feedback on code blocks, understanding concepts, where my code went wrong etc. Asking it all of the dumb questions we go trawling for.
But I can't speak to how this translates when you're a more intermediate developer.
I... assume that was meant sarcastically, but it's not at all clear from context I think.
That said, I still try to figure out the logic myself first, then let AI help polish or improve it. It is a bit slower, but when something breaks, at least I know why.
AI has definitely lowered the barrier. But whether you can actually walk through the door still depends on you.
The open questions right now are how much of a demand is there for more software, and where do AI capabilities plateau.
But, there is a key distinction that we would be remiss to not take note of: By definition, farmers are the owners of the business. Most software developers aren't owners, just lowly employees. If history is to repeat, it is likely that, as usual, the owners are those who will prosper from the advancement.
fruits and all non-essential food items are famously very elastic, and constitute large share of the spending.
for example: if cheap cereal becomes abundant, it is only at the cost of poor quality, so demand for high quality cereal will increase.
the LLM driven software engineering will continuously increase the bar for quality and demand for high quality software
In the long term, food demand is elastic in that populations tend to grow.
The more awkward truth is that most of what developers have been paid to do in the 21st century was, from the larger perspective, wasted. We mostly spent a lot of developer time in harvesting attention, not in actually making anything truly useful.
Most organizations do derive net benefit from laying off the below average and hiring the above average for a given compensation range, as long as the turnover is not too high.
And this delta increases when the above average can augment themselves more effectively, so it seems we should expect an even more intense sorting.
However, I do agree that the premium shifts from mere "coding" ability -- we already had a big look into this with the offshoring wave two decades ago -- to domain expertise, comprehension of the business logic, ability to translate fluidly between different kinds of technical and nontechnical stakeholders, and original problem-solving ability.
So much room left. As I doubt every developer will double check things every time by asking.
I'm also a bit tired of running into people that are 'starting a contracting firm' and have 0 clients or direction yet and just want to waste your time.
And if I were to jump into instruction-level programming today I would start by asking an LLM where to begin...
I browse the web. Eventually, I review the agent code and more often than not, I rewrite it.
Damn straight we are.
Just, don’t skip out on learning the fundamentals. There’s no royal road to knowledge and skill. No shortcuts. No speed running, downloading kung fu, no passing go.
Why?
Because the only thing LLMs do is hallucinate. Often what they generate is what you’re looking for. It’s the right answer!
But if you don’t know what and L1 cache is or how to lay out data for SIMD; no amount of yelling at the bot is going to fix the poor performance, the security errors, and the logic errors. If you don’t know what to ask you won’t know what you’re looking at. And you won’t know how to fix it.
So just remember to learn the fundamentals while you’re out there herding the combine space harvesters… or whatever it is kids do these days.
There actually is a ChadGPT but I assume the OP meant ChatGPT
I heard that before. Borland Delphi, Microsoft FrontPage, Macromedia Flash and so on. I learned how in 5 years or so, these new technologies would dominate everything.
Then I learned that two scenarios exist. One of them is "being replaced by a tool", the other is "being orphaned by a tool". You need to be prepared for both.
That said, even if the specific products like Cursor or ChatGPT are not here in 5 years, I am confident we are not going to collectively dismiss the utility of LLMs.
Heck I'm so tired of statements like this, many who? It's already a lot an LLM that automate/help the boring/tedious part of my job, I have yet to see taking over 2, 5 or 10 of my collegues, just knowing what a hawful lot these tiredlessly dudes do I couldn't ever imagine doing also their job. imo such statements have very short shelf life