[mercedised]

IMO, untill AI is intelligent enough to understand sarcasm it cannot replace a human. The title should be the golden era of Indian software engineers might be over. There are cheaper alternatives in the form of Phillipines, Indonesia with young populations emerging. One of my friend was working in a IT company which designed custom solutions for Hotels and Industries, the company consisted of 300 employees, post-covid they slowly started firing employees and as of now there are 15 regular employees in the company, all the work the IT professionals used to do has been outsourced to Phillipines due to cheaper costs. I am not a IT guy but the arbitrage cost advantage that India had is now over and that is more of a threat than AI. Experts plz correct me if I am wrong here.

One more threat is our education system, we all know how much gap there is between what a industry requires and what is being taught. We are no match with US and China in terms of AI developments, that is also one of the major threats to our future IT engineers. Our education system is not industry ready, has never been and in todays rapidly evolving technology we dont know if we will be able to keep up.

I would like to add one more story which is a true incident :

One of my friends who did IT engineering from a very ordinary college, and he was a very ordinary student. His ancestral business is dairy farming since ages. In the wave of everyone wanna become an engineer he also completed his engineering around 2011. He did not get a job and started doing dairy farming, his younger brother also did become a CS engineer could not secure a job and eventually started sitting on his ancestral dairy shop. Both of them with poor knowledge wanted to get a job but did not get it. After around a year in dairy farming business, my friend who had some other friend in a local IT company , he asked my friend if he would like to join his employer as an IT engineer for Rs 15000 pm. My friend who was just holding his life somehow, did a 3 months crash course, revised his engineering knowledge and joined the company in 2013. After a year of joining he trained his brother for 3 months and then his younger brother also joined the same company in 2014.

Tadaa, the friend of mine is on a 3 crore p.a. package in adobe and now resides in USA since 2019. His younger is still doing job in India on a package of 25 LPA. Both of them as I am very close to them were very very ordinary engineers with only degree being the proof that they have done engineering. I would have never employed them as a engineer if I was the employee. I have seen plenty of such examples where a mechanical, civil engineer did a 3-6 months crash course and now some of them are in US and most of them are earning a very healthy salary. This brings me to the question how easy it has become to be a software guy or a coder. Now, this is fortune for me. This is a joke I as a ECE engineer laugh at whenever life throws garbage at me.

[chucknorris]

Originally Posted by rkv_2401

Do you think making a MERN stack front-end, back end and database and hooking it all up together is all it takes to make a "fully-functioning clone" of WhatsApp or YouTube? ...

While I don't necessarily agree with the way rkv_2401 worded their post, they seem to be the only one (or one of the few, haven't gone through all the posts) that have a realistic take based in reality. This community is not very Software Dev heavy so I'm not surprised that people are more aware of the headlines and the flashy posts and less aware of the 100s of 1000s of videos, blogposts and articles by actual industry experts (one's without a leg in the AI sales game) that go over and with great technicality and provable facts that most of what is being thrown around by AI startups and VC's funding these startups is bullshit.

The only thing AI's will do in the field of software dev is act like glorious LSP's, if they are used for anything other than that (which is what's happening now with a lot of young people thinking that they can build million $ empires by having copilot generate code for them) it will lead to the need of millions of more skilled engineers to fix the crap that AI's generate. In fact it's already happening, there have been studies on code generated by AI's that shows that AI's are now training themselves on the shitty code spat out by other AI's leading to progressively shittier code. And this cycle will continue without a skilled dev that tells then "No, this is wrong".

I highly recommend ThePrimeagen and NeetCode on YT for people who care to look for more information and not just puff the AI joint being passed around.

[DigitalOne]

The AI Hype is dying.

https://www.youtube.com/watch?v=3A-gqHJ1ENI

[Samurai]

Originally Posted by DigitalOne

The AI Hype is dying.

The bugs are going up? I am not surprised.

Originally Posted by Samurai

Most IT work output is deterministic, and not probabilistic. No banking customer will be happy with a banking software that is accurate 99% of the time. If your salary is not deposited, would you be satisfied with the fact that 99% got their salaries? An IT customer would want their features work exactly as required, every time. You can't hide behind AI when customer yells at you.

As I pointed out last year, if you let AI generate code, who is going to debug the tiny 1% errors it will inject somewhere?

It is great for creating boiler plate code. But you need to read every line and understand what it is doing. But if you are staring 1000s of lines of generated code, who can do the reading and fixing. No one...

[V.Narayan]

Google CEO Sundar Pichai, rings alarm for software engineers, at the Q3 2024 earnings call and said that over a quarter of the company's codes are now written by artificial intelligence (AI).

https://www.indiatoday.in/technology...519-2024-10-30

Google CEO Sundar Pichai revealed a striking statistic during the company's recent third-quarter 2024 earnings call: over 25 per cent of Google’s new code is now generated by artificial intelligence (AI) and subsequently reviewed by human engineers. What does this mean for software engineers and coders? It marks a fundamental shift in the coding landscape, where AI increasingly shares the workload. It may not necessarily mean coders losing their jobs, but that AI can enable engineers to focus on higher-level problem-solving and innovation. That said, while automation boosts efficiency, it also raises questions about the future need for entry-level and routine coding jobs, pushing engineers to adapt their skills to remain competitive in an AI-driven landscape. This means, coders and software engineers may need to develop complementary skills in overseeing, refining, and guiding AI-generated code.

Some of us on team BHP including @Samurai & I have been pointing this out for over a year that the era of lucrative and growing software jobs in India is drawing to a close. AI is providing companies i.e. the employer a route out of the perennial high attrition rates in the software industry, the constant wage bargaining, constant threat of ‘pay me this or I exit’ situation. Of course Indian software companies share their full part of the blame for this sorry unstable scenario with their borderline unethical practices. It isn’t as if only young aspirational employees are to blame.

Without wasting time on the blame game AI is genuinely disrupting the scene. Not good news for junior young employees entering the workforce. Many on this thread debunk the notion that AI will ever be much more than the entry level boiler plate coding tool. I don't see that in the little exposure I have had. I see large companies going full throttle for AI to develop it to its full potential and reduce their dependence on employees. As Samurai writes in his post # 79 humans will be needed to clean up and tighten the last 1% i.e. the last mile.

[Samurai]

over 25 per cent of Google’s new code is now generated by artificial intelligence (AI) and subsequently reviewed by human engineers. What does this mean for software engineers and coders?

I will still raise the same question. Can programmers review generated code beyond a certain size? If all code gets generated, eventually no one will know how to write code. Can such people review code at all?

Google is known for its blunders in technology. Most importantly google is not into customized software services, which is what employs 80% of the IT industry. Building customized software for customers using AI is a very different ballgame.

[warrioraks]

Originally Posted by V.Narayan

Google CEO Sundar Pichai, rings alarm for software engineers, at the Q3 2024 earnings call and said that over a quarter of the company's codes are now written by artificial intelligence (AI).

AI is a real productivity booster. The usually touted aspirational goal in the industry is to enable 10X programmers. i.e. coders who are 10 times more effective because of AI tools. Even if this goal is achieved (which is a long shot), it will not replace the programmers. Ask any business or product team and there is always too much to do.
From a practical standpoint, AI enablement is a slow grind which will take many years to reach a steady state. The evolution will be similar to automation in car factories. It will be incremental in nature but once a part of the assembly line is streamlined using machines, there will no looking back.

PS: It would be interesting to observe Google's Profit per Employee over next few quarters. Ideally it should shoot up inorganically considering these claims.

[am1m]

Originally Posted by warrioraks

AI is a real productivity booster.

Seeing this happen even in my non-core-programming job function. The same job that I've worked as a part of a team of 4 people in the past, now I get done on my own. But more than immediate replacement, I've seen it happen gradually. Rather than firing 3, companies just didn't hire extra. So like you already said, it might happen gradually. But can't really say though. Job security was never a given with any single company in the IT industry anyway. Now I think overall, the number of jobs is certainly going to drop each year. All in all, think I'm happier that I'm closer to the end of my career than at the start, with these developments.

[warrioraks]

Originally Posted by am1m

Seeing this happen even in my non-core-programming job function. The same job that I've worked as a part of a team of 4 people in the past, now I get done on my own. But more than immediate replacement, I've seen it happen gradually. Rather than firing 3, companies just didn't hire extra. So like you already said, it might happen gradually. But can't really say though. Job security was never a given with any single company in the IT industry anyway. Now I think overall, the number of jobs is certainly going to drop each year. All in all, think I'm happier that I'm closer to the end of my career than at the start, with these developments.

I will be mixing up a few things in this post. So bear with me.

Yes you are right, this trend is not limited to programming jobs. It is impacting all white collar jobs at a minimum. Also agree with your sentiment that jobs are going to come down. My guess is that is primarily because of slowdown in global economy rather than AI. Just look at the headline numbers - The global economic engine (USA) is growing at 2% despite all the money printing. European heavyweights like Germany and UK are barely staying above 0% growth. China is trying to avoid a deflation and a property crisis (the growth is at historic low).

How will AI really impact us is a tough question and really hard to predict the future. But if we look at last 50 years, there have been various times where technology was supposed to be a big disruptor to jobs but it just ended up being incrementally better over the years. Some of these events/points have been -
1. The popularity of ERPs in the 90s and 2000s. There was a wide belief that with these integrated systems, organisations will not need any other custom software for supply chain, CRM, etc. This has been proven very wrong.
2. The advent of accounting softwares and Excel. People thought accountants jobs are at risk and will soon vanish. Again did not hold true.
3. Internet and digital boom. Most experts thought brick and mortar will be usurped by these new age companies. More than couple of decades and still not happened yet.
4. RPA (Robotic Process Automation) around 2015 - I ended up in so many tech transformation conversations where the primary theory was that RPA and no code analytics combined will eliminate all entry level jobs completely. Almost 10 years now, and no one even thinks it's a threat to jobs anymore.
5. AI is the latest trend. I am a big believer in this technology to significantly reduce the burden of boring grunt work making everyone's life better. But at the end of the day, just because it can speak like humans does not mean it can achieve outcomes like humans. As an example, it's almost 2 years now that Image and Art Generation models like DALL-E were launched. How many organisations have fired their graphics or UI teams? Adobe, Figma and Canva have integrated GenAI capabilities within their workflows but we still need humans to prompt the system and design it end to end.
6. On software programming - I might have said it multiple times on the forum. Post Covid world has seen a flood of below average, overpaid coders joining the workforce at scale. I cannot think of any other industry where salaries have gone up but the expectations have gone down. It's time for a coorection. Either these programmers upskill or they will be replaced not by AI but by people who use AI effectively.

Lastly, I strongly feel that from a global economy standpoint, there is a dire need for a catalyst to shake up things. Just like internet has led to new jobs, revenue streams and tremendous growth in last two decades, the world is ripe for another disruptive force. AI seems to the leading contender for now.