Anonymous wrote:More than half the software engineer population will be redundant. No one is hiring new folks to replace the laid off employees. Why would a tech company hire a newbie when the market is flooded with experienced unemployed who are ready to work for far lesser than before. The outlook is pretty bleak unless you’re a genius.
I’ve been in IT for over 2 decades.

why would anyone keep on employees at 250k when you can hire people with more recent tech skills at 125k?

Anonymous wrote:AI will replace a lot of desk jobs, but it won't replace jobs that require hands and human interaction. Choose your major accordingly.

desk jobs? you mean people who work on computers? yeah. but you're right - the trades will be fine

AI will replace a lot of desk jobs, but it won't replace jobs that require hands and human interaction. Choose your major accordingly.

I work in water - or Big Water as I like to say - as an engineer and AI is def coming for this area. I hope so!

this interesting, but trades are full of things like this. google water management and AI and you can read all night. I'm sure there are loads of other areas where AI is already making inroads

https://www.unesco.org/en/articles/applications-artificial-intelligence-water-management

Anonymous wrote:AI doesn't even understand physics. See this
https://www.thealgorithmicbridge.com/p/harvard-and-mit-study-ai-models-are

It gets the physics completely wrong but gets the orbits right -- but people did that too in the distant past before they figured out Newtonian mechanics (epicycles). AI (or at least the current version of deep networks) optimize for prediction and matching, not concept abstraction. I keep on top of this research for my work, and AI is too far away from this. The situation is far worse in biology.

And before someone jumps in to say that it's only a matter of time, hardly anyone (in academic research or at companies) is optimizing for this. AI firms have bet on AGI, but most of those designs are just beefed up transformers -- which are powerful but have serious limitations.

Read the article. They made a tiny toy model and complained that it couldn't solve a problem that took the smartest person in human history to solve.

So, some AI today might not be able to discover gravity from scratch? Guess who else can't that! You, and and at least 7 Billion other people alive today.


They removed exactly the thing they said it would need -- world knowledge, and then said that it didn't work.

People in this thread don't understand what fine tuned powerful models can do, and think it's some gotcha that ChatGPT free model can't replace month's work of expert work with 5cents of computer time.

Meanwhile, AIs are replacing humans in written, picture, and video communications, discovering drugs, building and debugging software, and more every day.

Anonymous wrote:Until AI can accurately render a human hand, I'm not all that worried.

It's not 2022 anymore. Wake up, Rip van Winkle

Anonymous wrote:No one is getting on a plane designed by AI. No one is moving into a skyscraper designed by AI. No one is launching a rocket designed by AI. No one is buying a car created by AI. No one is relying on a power plant built by AI. And on and on. There is a nexus of creativity and technical ingenuity in engineering that cannot be replicated by AI. And if AI ever gets to the point where it displaces engineers, civilization will have already collapsed long ago.

People are overestimating where AI is presently. Look at what Google has been doing recently. Your searches are getting AI responses, which can be helpful for simple things. But it always misses nuance, context and complexity. AI is making us dumber by the minute. But engineering can't afford to be dumb.

Would you rather get into a plan designed by a human engineer ast Boeing, and have pieces fall off mid flight?

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:there's a steady drumbeat of "AI will replace coders" and "AI will replace lawyers"

and yet everyone is still pushing their kids into engineering. I dont get it. engineering seems as/more vulnerable to me.

Engineers have to take into account human behavior as well as physics. AI just doesn’t understand human behavior.

So...engineers dabble in psychology too? Yeah, right.

Google AI summarizes:

"Engineering psychology, also known as human factors, is a field that studies how people interact with technology and systems, aiming to improve user experience, safety, and performance. It applies psychological principles to design and evaluate products, systems, and environments. Essentially, engineering psychologists work to make technology more user-friendly and efficient by understanding human capabilities and limitations."

More than half the software engineer population will be redundant. No one is hiring new folks to replace the laid off employees. Why would a tech company hire a newbie when the market is flooded with experienced unemployed who are ready to work for far lesser than before. The outlook is pretty bleak unless you’re a genius.
I’ve been in IT for over 2 decades.

Anonymous wrote:there's a steady drumbeat of "AI will replace coders" and "AI will replace lawyers"

and yet everyone is still pushing their kids into engineering. I dont get it. engineering seems as/more vulnerable to me.

Then you do not understand what engineering is. It is less replaceable by AI than most

property setbacks seems like something it could *easily* do.

the issue isn't that AI is going to take all engineering jobs. it's that it will take half of them. mostly because it will take an engineering - using AI - half the time to do their current work load. So work loads will be adjusted.

And it won't stop there because 5 years from now, when AI is even better, cities will be using AI to make a first past on submitted plans. that's a good thing - no plans getting lost in a DOB circle for two months. And big companies will be using AI to look at better water system management and electrical system design. This stuff seems fully within reach - it will be more an issue of when cities will get around to implementing. The private sector in chemical, nuclear, and space will be take advantage more quickly.

TLDR: There's zero doubt AI will come for engineering.

Anonymous wrote:

Anonymous wrote:there's a steady drumbeat of "AI will replace coders" and "AI will replace lawyers"

and yet everyone is still pushing their kids into engineering. I dont get it. engineering seems as/more vulnerable to me.

Engineers have to take into account human behavior as well as physics. AI just doesn’t understand human behavior.

So...engineers dabble in psychology too? Yeah, right.

AI is a “mostly right” engine— the only way it replaces engineers is if PE etc decide that it’s ok to be “mostly right” about engineering. Not saying it won’t happen but we are screwed if it does.

Anonymous wrote:

Anonymous wrote:AI doesn't even understand physics. See this
https://www.thealgorithmicbridge.com/p/harvard-and-mit-study-ai-models-are

It gets the physics completely wrong but gets the orbits right -- but people did that too in the distant past before they figured out Newtonian mechanics (epicycles). AI (or at least the current version of deep networks) optimize for prediction and matching, not concept abstraction. I keep on top of this research for my work, and AI is too far away from this. The situation is far worse in biology.

And before someone jumps in to say that it's only a matter of time, hardly anyone (in academic research or at companies) is optimizing for this. AI firms have bet on AGI, but most of those designs are just beefed up transformers -- which are powerful but have serious limitations.

give it 18 more months.

Np. You have no idea what you are talking about. The PP is absolutely correct. If all of physics is simply about optimization, maybe. But to discover the underlying physics, no.

Anonymous wrote:

Anonymous wrote:I’m an engineer. AI can’t do what I do, i.e., creative thought.

Yeah, you can stay for the creative thought—for now—and it can do the rest. You might keep your job, but all the entry level people won't. Same with architecture, but even that is going to go when you can eventually tell an AI, design me a classic new england saltbox for this plot of land—it will visualize it and then do the schematics in a few seconds.

There have been stories about how AI will replace architecture but nothing really on the horizon except pretty rendering companies.
How do you tell AI about property setbacks?
Balancing FAR with the actual flow of rooms and the constant back and forth with clients about the layout they have to live with? Especially because so many clients are obsessed with their window views?
Most projects require permitting, can AI acquire an architect or engineer's license for stamping and filing? Do insurance companies provide professional liability for AI?
Is AI going to manage the plumber, structural engineer and mechanical engineer to avoid clashes in their systems? There is industry software that sort of does that but every consultant has to use it and pay very high fees for the privilege. And at the end of the day, the clash detection is done by a human at any of these firms.