[masterChief007]

All repetitive pattern based work are the first things to get replaced by AI, the technology is good for that. Engineers who are dealing with algorithms and gdb/state machine based debugging are for now safe (but won't be in the long run, as these things are not that easy to feed to an AI system, but where there is a will there is a way).

But creativity is difficult to define based on patterns (yes sure the 1:3 fine arts related basic rules, but it still won't achieve the brilliance of the human mind)

E.g:
Maybe I am shooting at my own foot by saying these debuggers can be rigged to provide such debugging data for a similar sort of problems and what all steps are being used by the human engineer to fix that specific problem can be tabularized. Then one day lo and behold !! Automatic big fixing is on and no more need of maintenance teams !!

I have seen systems learning a language semantics and generate automatic codes on its own, done as a weekend trial (well maybe not a weekend, but you get the gist) by a very brilliant senior in our organization (and it was his hobby project).

[bsdbsd]

Originally Posted by JediKnight

Cognizant to lay off 3,500 employees. It will also give up 11 million sq feet in office space, or 80,000 seats, in large cities.

Source

If you know anyone who is 13+ years experienced, technically hands-on, and interested in a architect or a senior technical position, please contact me privately on my handle.

[Tucker48]

I'm happy that AI is making dumb jobs a thing of the past. BUT, it's not AI, it's automation.

Understand the difference between AI and Automation. Don't fall for news/analysis etc. How high would you rate the skill of driving amongst all skills? And still they failed because they are trying to automate an environment with so many variables.

Intelligence just can't be based on YES/NO (0/1). Digital computing will always have this problem.

Please don't fall for these reports which say AI is going to take over jobs. It's the automation which is taking over jobs since decades. Automation is a closed loop intelligence, means works only in a specific set of variables.

Also, there is cost to benefit ratio wherein some sectors, even automation fails. Talk about cheap labour.

[NewUser123]

Most of the articles and opinions I read on AI are mostly by people who have not actually played with it.

Here is an article by someone who plays with it and writes about how he did it, all in just 30 minutes…

https://www.oneusefulthing.org/p/sup...-in-30-minutes

I am concerned for entry and mid level employees in nearly all services fields.

[warrioraks]

Originally Posted by V.Narayan

At the most optimistic end AI might turn out like the dread of computerization was in the 1980s and 1990s where we feared jobs will be swallowed up but in reality, they made all of us more efficient and productive and in the end created more jobs both direct and indirect. But in 1985 or even 1995 we didn't know that. At the worst end AI could be a real job destroyer.

Thanks for sharing about the conversation. I know of a company which has given an entire department a 12 month notice. The employees of the department are expected to find a job outside organization since their entire process is getting automated that will require not more than 10% workforce to sustain current level of operations.

The learning capabilities of machines have grown leaps and bounds over last 5 years. If companies are given a free reign, I expect a lot of jobs being lost in the immediate future. Especially in large enterprises where processes are streamlined and division of labour ensures compartmentalization of tasks easier to automate.

I would hazard a guess here. Just like industrial revolution changed the manufacturing landscape and completely eliminated a lot of manual jobs. The AIML revolution will do the same for services sector over the next 10 years. The major brunt will be borne by people working in so called large IT factories (read IT, ITES, product companies). Hopefully some other sector will take it's place because our world is not yet to ready for a scenario where humans don't have jobs.


PS: Some folks have commented that AIML can only replace jobs that deal with following a pattern of repeatable tasks/activites. I think the technology has gone way past that now. Just explore DALL-E-2 capabilities and you will know what I mean. The system can throw out an entire collage of creative art faster than I typed this post.

[wheelnpaddock]

IBM to replace 7800 jobs with AI pauses hiring for those positions:

https://arstechnica.com/information-...ced-by-ai/amp/

IT has a trend of putting old wine in a new bottle, adding a few more bells and whistles and renaming it.

Computers are not different from AI. They represent early days of AI. The oldé calculator is in itself the most basic ‘artificial intelligence’. AIML represents a more evolved / advanced form of that intelligence.

This is Arthur C Clarke’s prediction in 1964:

https://youtu.be/YwELr8ir9qM

Short extract of the video predicting the end of biological evolution and the rise of machines. In 1964, it would have appeared as nonsense. For most it would still do. But given the accuracy of his predictions, it seems closer to reality than it did in 1964.

https://youtu.be/xkhddj7Alic

[jerrindanny]

Originally Posted by NewUser123

Most of the articles and opinions I read on AI are mostly by people who have not actually played with it.

Here is an article by someone who plays with it and writes about how he did it, all in just 30 minutes…

https://www.oneusefulthing.org/p/sup...-in-30-minutes

I am concerned for entry and mid-level employees in nearly all services fields.

I went through the above-mentioned article. The author clearly has covered the roles of both entry and mid-level employees of IT and marketing domains. But as big MNCs are now restricting the use of AI tools within their network, the ones that are going to reap the most benefit are startups and small-level organizations.
Though people who are looking at freelancing are going to reap the fruit.

[NewUser123]

Originally Posted by jerrindanny

But as big MNCs are now restricting the use of AI tools within their network, the ones that are going to reap the most benefit are startups and small-level organizations.
Though people who are looking at freelancing are going to reap the fruit.

Thanks for the feedback, but as per his latest post..
“ Microsoft is planning on adding these capabilities to every single Office program in the near future, so mass adoption of AI for work is going to be happening much sooner than you think. Expect automatic creation of Word documents, automated Excel analysis, PowerPoint created with a paragraph of text (with images generated by DALL-E), and a Microsoft Teams that send you to-dos and suggestions for improvement after every meeting. This is not science fiction. All of these things are already announced.”

https://www.oneusefulthing.org/p/it-...to-get-strange

I feel all the gains of startups and freelancers might end up being temporary.

[AirbusCapt]

My take on this automation/AI taking over jobs discussion:

Machine learning and artifical intelligence will take over all repetitive and predictable jobs. If your job involves testing code, believe me it will be the first to be completely automated. If your job involves making usecases for testing, that job will survive since a machine cannot think outside the box like a human.

Likewise a funtional expert or a software architect function will still be human powered, however the whole "coding" part of the software process will be automated. Therefore; the "manager" function will also be redundant.

As a software person if you want to be employed think of what jobs/roles that involve real thinking and creative though process and reskill yourself for those roles.

[NiInJa]

AI is going to take away low level jobs in IT, that has already started. Basic coding, testing and automation type of work is handed over to machines. People will have to think beyond tasks that machines cannot handle or do. It just means humans will have to become smarter with their work choices.

I see a change in approach when we take interviews of freshers or less experienced folks these days. Gone are the days when we used to ask freshers to write codes during initial assessment, a simple search on google then and ChatGPT now can spew it out in seconds. The focus on interviews is now is more on how clearly can candidates define a problem and how efficiently they can design your solution and how expressive they are .
In our field which focuses on Financial Planning solutions, AI is still in a nascent stage and might take years to take over as it still very 'people' centric. It is not just the math and science, but politics, negotiation and presentation skills, human relations etc that add the numbers

Have an interesting story to tell.
Back in 2017, one of my college friend was talking about AI and how its going to disrupt everything while we were lying on a beach, staring at the sky. He told me that I need to learn about this urgently and questioned my responsibility towards the society as a 'computer engineer'. It was late night, I was inebriated and it got me angry and pumped. I declared that day that I am going to learn ML right here right now. Downloaded Udemy on my phone, bought an online course in Python in the heat of the moment, at Kashid Beach (Alibag), and in three months, learnt basics on ML in Python. I used this learning to demo a very simple, but effective use of ML in finance to my office as a part of marketing and promotion material. When I presented my idea to the folks, my then manager asked me 'kyu kar raha hai ye sab ?' I could not defend myself effectively and we dropped it and instead focused on some stupid Groovy language training sessions. Today, the same thing that I created back in mid 2017 is marketed by Oracle solutions under the title 'BYOML' (Bring your own ML) while one can get Groovy scripts generated by ChatGPT !

[efgritesh]

Originally Posted by NiInJa

I used this learning to demo a very simple, but effective use of ML in finance to my office as a part of marketing and promotion material. When I presented my idea to the folks, my then manager asked me 'kyu kar raha hai ye sab ?' I could not defend myself effectively and we dropped it and instead focused on some stupid Groovy language training sessions.

This has been the biggest tragedy of my life. I have a keen eye to identify broken things and come up with simple solutions for it. Unfortunately most of the time, idea would be difficult to sell as the managers are often too occupied with their own set of problems, and don't want additional items to bother them.

not sure if AI can be used to quantify potential of ideas/solutions in future

[AltoLXI]

Shraddha's take on performance appraisals.

https://www.youtube.com/watch?v=fZYlNky6AsM

[mallumowgli]

Belongs to this thread, perhaps?

https://www.economist.com/finance-an...l-intelligence

[rajshenoy]

As someone who has spent couple of decades in this industry, AI we have seen so far is just the tip of the iceberg.

For years, we have been focusing mainly on automation. Software itself is all about automating repetitive work. However due to the shortage of skilled labor, we always had short supply. Also most of the IT jobs in India are only service focused so entire roadmap and the demand is mostly driven by so called "masters" sitting overseas.

Let's face it, western economies are not doing great and will continue to struggle in coming years. On the other extreme, thanks to herd mentality of parents, we have more engineering colleges on every street now with very average talent passing out. Focus is very skewed towards this IT services industries as against specialised/manufacturing industry.

So the result we are seeing today are due to these events and not really due to AI.

We are increasingly seeing AI at work in chat bots(I would say still primitive form of AI, more on conversational abilities only) and technical support automation. This definitely would have impacted low paying support jobs and will continue to do so as AI becomes more powerful.

I consider myself lucky to get opportunities to continuously upskill my resume and currently working on Generative AI based models. This sometimes amuses me and also equally scares me as this AI can potentially render lakhs of our average engineers jobless. I sometimes jokingly say the next generation may not have to ever write code.

We definitely need our universities to focus on quality of talent than sheer quantity.

Here is my take on next decade with AI - we will see more applications of AI with human like creativity and emotions. This I feel will have impact across higher spheres of life, such as creative writing, customer support and eventually AI powered softwares (beginning to see these). We are still long way from seeing future offerings such as self driving cars, autonomous robots etc. However we do have their primitive applications already in front of us (ADAS and industrial robots etc). Having said that, AI is here to stay and will only become more powerful. I may not see terminator movie series in my lifetime though

[Samurai]

Originally Posted by rajshenoy

I consider myself lucky to get opportunities to continuously upskill my resume and currently working on Generative AI based models. This sometimes amuses me and also equally scares me as this AI can potentially render lakhs of our average engineers jobless. I sometimes jokingly say the next generation may not have to ever write code.

I used to be alarmed at this once. In fact, I had given public speeches around 2015-2016 on how AI is going to take away most IT jobs. Then this fear made me dig deeper AI/ML, etc., to get ready for job loss tsunami. But the tsunami didn't come for next few years. Instead, jobs and wages went up during the pandemic. Data sciences related fields started creating more jobs, while reducing none in the traditional IT.

Which made me re-examine my analysis from few years ago. Now I am a software stack vendor to AI companies, therefore I have better understanding of what AI can and cannot do in the immediate future.

The main distinction is deterministic vs probabilistic. We know Generative AI is all a big rage nowadays. AI can create incredibly realistic images, essays, videos, etc. Yes, copywriters, artists, creative writers will be directly affected. A new image/writeup can be created in dozens of different ways and it will be acceptable. There is nothing like 100% accuracy when it comes to artistry. It is always in the eye of the beholder. Therefore, a probabilistic output would be fully acceptable.

However, most IT jobs are not competing with Generative AI. IT jobs are at most competing with Cognitive/Enterprise AI.

Is Cognitive AI mature enough to replace IT jobs yet? 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.

Recently I was talking to someone in an IT MNC who has been asked to generate 100% code using AI from the user requirements. We both were laughing at the insanity of it.

1) The user requirements has to be so thorough, it might as well be code itself. Who is going to write this? Surely not ChatGPT.
2) The AI may generate a million lines of code. No one can afford to review that code. Automating the code review will be an oxymoron, since AI is supposed to generate code that need not be reviewed. Regression is built into AI anyway. So, there is no need to review?
3) Therefore, code shall be verified only via testing. Who is going to write these test plans? Same AI that generated the source code, after reading the user requirements? If the source code and test plans were created by the same AI entity, why bother running the tests? AI obviously would have generated code that will pass the test plans generated by it. So, there is no need to test?

Now remember that AI is probabilistic. That means AI generated software would work 99% of the time. How do you determine which 1% it missed, considering you can neither review nor test the code?

The practice of generating of boilerplate code has been around for many decades. AI will just take that to the next level, that's all. Applications will get lot more richer and complex. And programmers can focus on deterministic stuff that AI still can't do, which is a lot.