Lord Posted December 31, 2025 Posted December 31, 2025 3 hours ago, coffee_rules said: Thanks . Pun was intended. I know STD stands for subscribers Trunk Dialing) It also stands for ,uh........ Super tall dinosaur coffee_rules 1
G_B_ Posted December 31, 2025 Posted December 31, 2025 18 hours ago, Lord said: Did Boeing fire you because of your medical stuff? I'm in the same boat looking for a job. Doing some basic AI engineering courses or Udemy and Coursera as it is not my field You work for rbi?
Lord Posted December 31, 2025 Posted December 31, 2025 3 minutes ago, G_B_ said: You work for rbi? not working rn
G_B_ Posted December 31, 2025 Posted December 31, 2025 2 hours ago, Ranvir said: The question for any company laying off staff due to AI is, who will actually buy your product if people don't have an income? Henry Ford is largely responsible for the 2 day weekend in the west. It gave workers an extra day to buy his cars. That's how businesses need to think. Hmm in the west around 25% to 30% of the population is retired or financially well enough to retire. This figure is rising as soceity ages. Japan that figure is 40%. They badly need AI to maintain living standards. If a single office can do the work of 3 then it really works out.
G_B_ Posted December 31, 2025 Posted December 31, 2025 11 minutes ago, Lord said: not working rn Ok. Lord 1
coffee_rules Posted December 31, 2025 Posted December 31, 2025 58 minutes ago, G_B_ said: Hmm in the west around 25% to 30% of the population is retired or financially well enough to retire. This figure is rising as soceity ages. Japan that figure is 40%. They badly need AI to maintain living standards. If a single office can do the work of 3 then it really works out. True, most need a job only for medical insurance. Insurance is a killer, especially if you have some illness.
G_B_ Posted December 31, 2025 Posted December 31, 2025 31 minutes ago, coffee_rules said: True, most need a job only for medical insurance. Insurance is a killer, especially if you have some illness. See i think ai + robotics is the killer There are already a lot of dark factories emerging. Was just reading in the ft 10% of jobs in risk and compliance will go in finance. So mostly they will push out the over 50s by 2030 to make ot cost effevtive. In tech its the junior roles vanishing. Will target each indistry dofferently.
bharathh Posted January 1 Posted January 1 On 12/31/2025 at 8:37 AM, coffee_rules said: Can you suggest a few sources to learn? Need to get a job Depends on what you want to do (I know it's get a job - but what kind of job?). Just doing courses is not useful. Happy to help in any way I can. Please DM me if you want.
velu Posted January 1 Posted January 1 5 hours ago, diga said: @velu san, how are you managing your trading with AI? kya AI only use of AI is accumulating information abt stocks without much search we are old fashioned non-directional index option sellers
MK55 Posted January 1 Posted January 1 (edited) AI is going to reduce the workforce sizes in every field. The ones that will survive are occupations that rely more on judgement and the ones where the stakes are very high (such as life / legal etc.) So surgery fields (robotics will help but it will be powered or at least supervised / adjusted dynamically by a human). Also legally that human will be responsible. Medical diagnostic fields / radiology are much less safe compared to surgery. And no robotics is still nowhere close to supplant complicated surgeries. Trading is an good example. U can’t have every fund / desk running the same model. No one will make money as there will be no trades. So what you will find is traders / quant guys using AI to tune their own respective strategies (But less traders / researchers will be required). In fact no one uses AI directly. But in future it will be used to do more complex stuff like back testing , dynamic scheduling of tasks etc. Presently it’s mostly used in small automations / adhoc research code, trying to summarize observations, checking if you’ve missed something etc. - all still useful n time saving. Investment banking (except the few front line guys), risk jobs in banks are absolutely gonna be dead (large reductions). Risk in buy side is different as that’s linked with algorithms and not compliance BS. Consulting jobs will be and should be dead. McKinsey already laying off some now. It’s because they just evaluate n propose options. They don’t decide what to choose. Plus hugely overpriced. Agriculture (for quite some years) , human services (cooking etc) will be safe too. But certain types such as cleaning related, domestic transport related jobs, store cashier related jobs etc. etc. will be replaced. Influencer jobs will also be safe lol as they appear to build a human connection. But there’s competition. Edited January 1 by MK55 Lord 1
MK55 Posted January 1 Posted January 1 (edited) Also doing AI courses is useless - personal opinion. I would rather build core technical skills, stronger domain knowledge and then get involved in a job where I have to use my judgement to decide between solutions - even better if the choice has a legal implication and if it’s design / creative choices driven. May be a job where no two people can have the same answer (by design) is also a good choice. Edited January 1 by MK55
vvvslaxman Posted January 1 Posted January 1 Simple programmers job is at stake. Let me give you a cricketing example. Just as an exercise for this thread i induldged myself in doing an analysis Best Indian player against 7 teams ( minus minnows) in bilateral wins. So you have several criterias. You have runs, averages, strike rates, wickets, bowling average, bowling strike rate and eras. Basically originally i gave weightage to bowling. But it severaly punished batsmen. Then moved to 50 to 50. Then decided a simple logic of pre 2010 bowlers get less weightage, batsmen get more weightage. Post 2010 bowlers get more weightage, batsmen get less weightage. All this iteration and calculation took me 15 minutes to generate this table
bharathh Posted January 2 Posted January 2 10 hours ago, vvvslaxman said: Simple programmers job is at stake. Let me give you a cricketing example. Just as an exercise for this thread i induldged myself in doing an analysis Best Indian player against 7 teams ( minus minnows) in bilateral wins. So you have several criterias. You have runs, averages, strike rates, wickets, bowling average, bowling strike rate and eras. Basically originally i gave weightage to bowling. But it severaly punished batsmen. Then moved to 50 to 50. Then decided a simple logic of pre 2010 bowlers get less weightage, batsmen get more weightage. Post 2010 bowlers get more weightage, batsmen get less weightage. All this iteration and calculation took me 15 minutes to generate this table How do you know if the values shown are correct though?
vvvslaxman Posted January 2 Posted January 2 (edited) 27 minutes ago, bharathh said: How do you know if the values shown are correct though? We interact throughout this. It walks me through each step of calculation and prompt for confoirmation. You can also see "look at my thinking" or something . You can see the python code that is being run. sorry you mean the actual values? I cut and paste from cricinfo to excel and upload to the chat. Edited January 2 by vvvslaxman
MechEng Posted January 2 Posted January 2 On 12/30/2025 at 4:31 PM, goose said: my facebook feed is stuffed with shots of hardik panday cuddling ananya panday on a yacht. Just cuddling? Maybe in daytime only. At night, Hardik will go boom boom like a tracer bullet from both over the wicket and around the wicket.
raki05 Posted January 2 Posted January 2 But roles like AI engineer and MLops engineer will flourish-like building more and more AI enabled applications, build and deploy mlops model, creating AI agents or AgenticAI . Also adaptation go complete end to end ai wont be that quick and easy. Every role be it consultant, product manager to developer will require AI upskilling. Lord 1
MK55 Posted January 2 Posted January 2 4 hours ago, raki05 said: But roles like AI engineer and MLops engineer will flourish-like building more and more AI enabled applications, build and deploy mlops model, creating AI agents or AgenticAI . Also adaptation go complete end to end ai wont be that quick and easy. Every role be it consultant, product manager to developer will require AI upskilling. But wouldn’t those roles - “AI” engineer also diminish with time? Even now using agents to do some well defined or well specd problem isn’t anything complex. The true value add of that employee is to figure out what to specify. Which model to choose, how to break down and model the problem, which stochastic states to consider, parameters to include for those sub models, judge what looks spurious, designing the back testing scenarios n cohorts, which stress scenarios to even run, how to refine the assumptions based on your observations etc. There’s no one right wrong answer here. All this is just statistics, math, design choice, and domain knowledge. The ops part friction (related to AI) will keep getting lower n lower as AI matures.
raki05 Posted January 2 Posted January 2 1 hour ago, MK55 said: But wouldn’t those roles - “AI” engineer also diminish with time? Even now using agents to do some well defined or well specd problem isn’t anything complex. The true value add of that employee is to figure out what to specify. Which model to choose, how to break down and model the problem, which stochastic states to consider, parameters to include for those sub models, judge what looks spurious, designing the back testing scenarios n cohorts, which stress scenarios to even run, how to refine the assumptions based on your observations etc. There’s no one right wrong answer here. All this is just statistics, math, design choice, and domain knowledge. The ops part friction (related to AI) will keep getting lower n lower as AI matures. All those new models or submodels , stochastic states, adding new hyperparameter are the role of researcher and you cant get such roles until you are practicing researcher like Andrew Ng of stanford who also worked witg facebook and google or PHD in mathematics. There are such roles as well but getting those would be really tough. What i am talking is a general role in Manufacturing, Finance and Healthcare company which you can target as IT professional. Where our main role would be either to expansive and often focuses precisely on integrating Large Language Models (LLM) into applications to create functional, real-world solutions or deploy already existing models with slight tuning based on your org specific needs within hour orgainisation thats what mlops engineer does like bridging data science and IT operations, automating and managing the lifecycle of machine learning models from development to production.Similarly there will be role ofAI/ML architect, Product Owner , Product Manager and so on. Building AI agents or creating agentic AI with cluster of agents and integrating them with various internal systems. Also integrating your LLMS with external systems using MCP protocol leveraging RAG to enhance contextual understanding of existing LLMs going to be a more creamy jobs. But all tgese adaptation will take it times and we should be hopeful to transition in to such roles with upskilling ouselves. What Elon Musk or Nvidia guy is saying is creating bubbles to raise their stocks the big change they are talking about where everything will be done by machine itslef is a bit distant dream. As AI can be bias depends in data it cant be ethical , inability to grasp nuance and context. Hence i am just focussingbon to uplift myself right now. BTW AI is not new as you know first perceptron or most basic neural network was setup in 1950s i believe its only creating wave dur to cloud computing. Next big thing i am more focused on is quants computing as that would be biggest shift in computer industry and it will change all ai models and modern algorithms like CNN, RNN.
Recommended Posts