AI Doom Predictions Are Overhyped | Why Programmers Aren’t Going Anywhere - Uncle Bob's take

Uh, I think I think the people that are

um predicting doom are just in it for

clickbait. Um, they like the idea of

predicting doom. It's going to be

terrible. Oh my god, life is going to

end. No, it's not. Um, there's other

another group of people that are saying,

uh, there will be no more programmers

because the AIs will do all the

programming. Let me tell you a little

story about that. The very first

programmers to worry that they were

going to lose their jobs because of a

technology improvement were programmers

in the very early 1950s.

These people were writing in binary,

literally binary. They wrote their code

in binary. There were no compilers.

There were no assemblers. In the worst

case, they actually got pieces of paper

tape and they punched holes in the paper

tape one at a time. They were the

programmers. They wrote code. They

called it coding because the holes in

the paper was the code. And Grace Hopper

came along and came up with an idea for

taking

a much

wasn't much a slightly better

representation, still numbers, not quite

binary anymore, but still numbers. and

automatically using the computer to

automatically translate that to the

holes in the paper tape.

And the programmers at the time said,

"Oh my god, anybody will be able to do

this. Then we're going to lose our

jobs."

This is what AI is. AI is another tool.

It's um a fairly useful tool as long as

you don't believe too much of what they

say. Um Chat GPT is perfectly willing to

lie to you and and tell you all kinds of

misinformation. All of them will. Uh if

you ask them to write code, they might.

They might write some code for you, but

you better check that code really well

because those those things have no idea

what they're doing. They're not

thinking, right? They are still just

programs that are assembling data

statistically to satisfy an algorithm.

They are not judgment. They have no

judgment. Uh as time goes by, I think

these these programs will get better and

better. AIS will get better and better.

That's good. We they will not approach

human intelligence. That is not on the

horizon. And I know people say, "Well,

we're only a year away from AGI." Yes,

we are. We are well more than a year

away. I I doubt that the current

siliconbased technology has the capacity

to emulate a human brain. I'm not sure

that's feasible. You know, we put down

our circuits on two-dimensional things.

We have not broken into the

three-dimensional neural net that you

have in your head. That's we're way away

from that kind of stuff. So, I don't I

don't think that's going to be an issue.

There will be some changes. There will

be some interesting things that happen.

Uh there will be great tools that will

help us, but the human will always be in

the loop. My prediction. Uncle Bob's

views on AI are becoming more and more

common as time goes by. And this is in

part due to the fact that we might be in

the thro of the solution stage of the

Gartner hype cycle. That stage with the

initial allure of a new technology

begins to fade. And people start

wondering if it can really deliver what

was promised. And this isn't the first

time something like this happened in

tech or in AI more specifically. We've

been here before in the 1960s and 70s

during the symbolic AI boom when logic

based systems failed to handle the real

world complexity and ambiguity of

language and perception. Again in the

1980s when expert systems once the

future of intelligence collapsed under

their inability to adapt despite massive

hyping newspapers and corporate

investments and once more in the early

2000s before deep learning triggered yet

another wave of optimism. This repeating

pattern is often called the AI cycle

when a breakthrough achieves what was

once deemed impossible leading people to

believe the artificial general

intelligence is just around the corner.

But after the initial excitement,

limitations start to show and the signs

are here again. Diminishing returns in

AI progress. Companies quietly scaling

back their AI hiring. Papers

highlighting the inadequacies of large

language models in real world settings.

Permanent thinkers in tech and AI

questioning whether current

architectures are even capable of

reaching AGI. all signs of a looming AI

winter. So, does this mean AI won't take

your job or should you still pivot to

something less risky? Well, when a

disruptive technology emerges, people

naturally start re-evaluating their

careers. And often that's a rational

move. Historically, if your field was

disrupted by automation, jumping ship

made sense. But software development

might be different because of the sheer

level of complex context dependent

problem solving it involves. From

interpreting human requirements and

navigating social context to designing

robust and maintainable systems to

completely replace a human developer, AI

would need to reach a level of reasoning

that might render most other occupations

obsolete as well. In complexity theory,

there's a principle that once a single

hard problem in a class of difficult

problems is solved. The rest tend to

follow. If AI ever reaches the point

where it can autonomously build

software, it might have also solved many

of the core challenges of general

intelligence itself. Meaning no

occupation would be safe from automation

in that way. And this ties into the

argument made in AI snake oil by Arun

Narayan and Sash Kapoor particularly in

their discussion of the latter of

generality where they note that you

can't meaningfully predict the next rung

on that letter meaning that each level

of AI capability brings new and

unpredictable consequence. So jumping

into another quote unquote safer

profession right now might be premature

or even a complete waste of time. The

social economic landscape could look

completely different by the time the

next wave of AI rolls through. So

investing time and money to pivot may

not protect you because the next

generation of AI could easily automate

whatever safer field you move to. Dots.

Subscribe for more.