ChatGPT AI Tool— We all work for robots now

[956 Worldwide 35598]

If you haven’t signed up for the ChatGPT AI tool, you’re missing out. It can do most of your job for you or it can do this:

 

Quote

Write three paragraphs explaining non-fungible tokens (NFTs) and whether they are a good investment in the voice of Nature Boy Ric Flair

Nature Boy Ric Flair here, and I want to talk to you about non-fungible tokens, or NFTs. These are digital tokens that are unique and cannot be replicated or exchanged for something else of equal value. They are often used to represent ownership of a digital asset, such as a piece of art or a collectible.

One of the biggest advantages of NFTs is that they provide a way for people to own and trade unique digital assets in a transparent and secure manner. This is because NFTs are built on blockchain technology, which allows for the tracking and verification of ownership and transactions.

Now, you may be wondering if NFTs are a good investment. In my opinion, it all depends on the individual. Some people have made a lot of money by buying and selling NFTs, but it's important to remember that the market for these tokens is still relatively new and unregulated. As with any investment, it's important to do your research and only invest what you can afford to lose.

Add "wooo" at least three times and the phrase "to be the man, you gotta beat the man"

Nature Boy Ric Flair here, and I want to talk to you about non-fungible tokens, or NFTs. These are digital tokens that are unique and cannot be replicated or exchanged for something else of equal value. Wooo! They are often used to represent ownership of a digital asset, such as a piece of art or a collectible. Wooo!

One of the biggest advantages of NFTs is that they provide a way for people to own and trade unique digital assets in a transparent and secure manner. This is because NFTs are built on blockchain technology, which allows for the tracking and verification of ownership and transactions. Wooo!

Now, you may be wondering if NFTs are a good investment. In my opinion, it all depends on the individual. Some people have made a lot of money by buying and selling NFTs, but it's important to remember that the market for these tokens is still relatively new and unregulated. As with any investment, it's important to do your research and only invest what you can afford to lose. But let me tell you, to be the man, you gotta beat the man! Wooo!

 

[52-80 19377]

[texashorne 1194]

10 minutes ago, 52-80 said:

My wife asks me that 2nd question all the time. 

[Nivek 17417]

requires my phone number? fuck that.

[Brian Fantana 6872]

32 minutes ago, Nivek said:

requires my phone number? fuck that.

You can just go to the website and sign in with Google or w/e

[Kennythetiger 9642]

1 hour ago, texashorne said:

My wife asks me that 2nd question all the time. 

“Was that on your website?”  

[texashorne 1194]

13 minutes ago, Kennythetiger said:

“Was that on your website?”  

We'll be on vacation and I'll suggest a restaurant and she'll ask "is that's what Surly likes? I'll respond yes, and they're also the reason we are on vacation at this place..

[Goofyboy 3434]

We'll be on vacation and I'll suggest a restaurant and she'll ask "is that's what Surly likes? I'll respond yes, and they're also the reason we are on vacation at this place..

Surly has not done us wrong on food recs. Wife asks me the same thing.

[HamsterHookah 112]

I haven't been nearly as impressed with this as the entire tech bro thought leader community, that's for sure.

[immamac 44387]

[Fuck Tim Beck 3904]

Trained professionals huh? Like South Austin's mom? 

[Guest]

I've read some stuff written by this. It comes across as very generic and like the sort of content that spammers write for SEO purposes.

[High Plains Drifter 11192]

2 hours ago, Brian Fantana said:

You can just go to the website and sign in with Google or w/e

 

signing in? Yeah fuck that.

 

[Captainant 28574]

4 minutes ago, FirstTimeCaller said:

I've read some stuff written by this. It comes across as very generic and like the sort of content that spammers write for SEO purposes.

That's basically all that NLP machines like this are lol. It's able to understand intent and reply accordingly, but it's still just parroting things that it's learned from its training datasets. 

When microsoft tried this back in 2015 using twitter posts as the training data, they ended up with a chatbot praising hitler lol. Garbage in, garbage out, and all that

[F250 29203]

It's pretty good with tech questions but doesn't know shit about sports.

[FartingMonk 3625]

Is this the same AI bot that told a 10 year old to plug in a charger halfway and put a penny on top when asked what sort of science experiment to do?

[HamsterHookah 112]

14 hours ago, FartingMonk said:

Is this the same AI bot that told a 10 year old to plug in a charger halfway and put a penny on top when asked what sort of science experiment to do?

No.

[HamsterHookah 112]

14 hours ago, Captainant said:

That's basically all that NLP machines like this are lol. It's able to understand intent and reply accordingly, but it's still just parroting things that it's learned from its training datasets. 

When microsoft tried this back in 2015 using twitter posts as the training data, they ended up with a chatbot praising hitler lol. Garbage in, garbage out, and all that

Yea, which is why (as you know, but for the benefit of the broader audience) there are ethical considerations and committees and teams dedicated to Big Data and Machine Learning, as depending on the inputs you can get biased or even racist "AI".

I haven't followed up on it but about a year ago there was a big controversy of Google firing their ML ethics leader or something and it allegedly being unethical.

[Mole 4344]

The songs it writes are hilarious.

 

Spoiler

Verse 1: George Washington was feeling quite ill

His stomach was aching and his bowels were moving still

He tried to hold it in but it was no use

The diarrhea was coming, he could not refuse

Chorus: Oh George Washington, our first leader true

With diarrhea running down his leg, what else was he to do?

He tried to hide it, but it was in vain

Poor George Washington, with diarrhea in his brain

Verse 2: He thought he could control it, but it was just too strong

The diarrhea kept coming, all day long

He tried to hide it from the world outside

But the smell was overpowering, and he could not hide

Chorus: Oh George Washington, our first leader true

With diarrhea running down his leg, what else was he to do?

He tried to hide it, but it was in vain

Poor George Washington, with diarrhea in his brain

Verse 3: But still, he persevered and he fought the good fight

Despite the pain and embarrassment, he held on tight

He knew that he had a job to do

So he put on a brave face and he carried on through

Chorus: Oh George Washington, our first leader true

With diarrhea running down his leg, what else was he to do?

He tried to hide it, but it was in vain

Poor George Washington, with diarrhea in his brain

Outro: So let us remember George Washington, and all that he endured

Despite his struggles, he remained strong and pure

His legacy lives on, and we will never forget

The time that George Washington had diarrhea and could not quit.

 

[ztejas 44236]

Is this the one that's writing people's code for them?

If so, I think a lot of you are being pretty naive. 

[Nice Guy Eddie 13740]

These new AI tools just tell me that some future jobs will be more about optimizing the usage of these tools. Or using them to create the base project and then modifying them as needed.

[956 Worldwide 35598]

8 minutes ago, Nice Guy Eddie said:

These new AI tools just tell me that some future jobs will be more about optimizing the usage of these tools. Or using them to create the base project and then modifying them as needed.

Yep. I’ve already seen someone who doesn’t use English as a first language write something fairly technical and ask it to turn it into fully correct English. It was amazingly good. 

[StruggleBus 3922]

This will die once they start charging for it

[fluff 87]

It’s pretty damn impressive if you have spent any time messing with it.

[Viking 301]

I was a skeptic but have come around.

For coding its great for simple bits but can't replace what most programmers actually do. I don't write 100 line apps but rather I write 100 lines of code in a 50k line codebase and do so without breaking any of the existing code. I doubt it'll ever be able to do that as it requires deep domain knowledge. Maybe it gets there but doesn't today.

I've sent it some of my google searches and the results it produces are a lot better. Eg most recent was "How long do I cook a quiche?". Google wants to send me to a bunch of quiche recipe sites which are all SEO filled garbage. ChatGPT just gave me the answer.

[Captainant 28574]

6 minutes ago, Viking said:

ChatGPT just gave me the answer.

Well, it gave you AN answer lol. No guarantee that'll be a good quiche 

Edited December 11, 2022 by Captainant

[SubliminalHorn 7009]

On 12/9/2022 at 10:28 AM, Goofyboy said:

Surly has not done us wrong on food recs. Wife asks me the same thing.

My husband is always telling me to get on and ask my surly friends their thoughts on __________

[Goredho 17662]

https://betterprogramming.pub/pair-programming-with-ai-writing-a-distributed-fault-tolerant-redis-client-using-chatgpt-239dc0a90550

The setup:

As a “Staff Engineer”, a part of your routine work is to find ways to make yourself and the team more productive. That includes implementing the right abstractions, adopting useful tools/frameworks and working with and training senior/junior engineers, and growing them to a point where they can function without handholding.

So how should one view ChatGPT? Should I think of it as a useful tool/framework? Or a smart junior engineer who just needs some direction? Let’s find out.

The payoff:

Given my experience, I believe that this can be used to speed up parts of software development; prototyping of simple components, building parts of a project in smaller modules, etc. It took me a little over 2 hours to get this working; I spent longer writing this post.

Giving it more context as I would to a real person seemed to help.

However, given the code quality issues and the fact that ChatGPT makes “creative” choices, means you have less control over things like how well the code performs. Code maintenance may also be harder.

Over time, as LLMs improve and are the result of training on much larger data sets, it may be the case that these “emergent properties” are more effective. Then could we envision these LLM’s reviewing their own code and likely writing better and better software, and even explaining how it works to us.

Learning to prompt these types of LLMs better could also be a very useful skill to develop and that may soon be a new job function. I’ll write up a future post on the techniques I learned to prompt ChatGPT better.

The rest of the article is full of low level code that might not mean much to most here.  What is  pretty interesting to me that being able to communicate with and direct the AI to get it to produce a working result for a non trivial coding project doesn't seem a whole lot different than the communication and direction someone would need to give a junior developer to produce a similar result.  I think there are a lot of people who are in line for a CS degree that are going to be disappointed when it's time to enter the workforce in 5 years.  In 10-15 years, maybe even the senior and above resources are threatened.  Brave new world.

Edited December 13, 2022 by Goredho

[Captainant 28574]

58 minutes ago, Goredho said:

The rest of the article is full of low level code that might not mean much to most here.  What is  pretty interesting to me that being able to communicate with and direct the AI to get it to produce a working result for a non trivial coding project doesn't seem a whole lot different than the communication and direction someone would need to give a junior developer to produce a similar result.  I think there are a lot of people who are in line for a CS degree that are going to be disappointed when it's time to enter the workforce in 5 years.  In 10-15 years, maybe even the senior and above resources are threatened.  Brave new world.

It's training a neural network either way between an ML model and a jr engineer - one lives in computers in a data center, the other is piloting a meat suit. 

[HamsterHookah 112]

On 12/11/2022 at 12:19 PM, Captainant said:

Well, it gave you AN answer lol. No guarantee that'll be a good quiche 

 

[956 Worldwide 35598]

3 minutes ago, HamsterHookah said:

 

I kept messing around with it and it is NOT good with writing styles. I asked it to do a Hemingway-styled paragraph and an e e cummings style poem and got piss-poor results. 
 

I also found I had to prompt it to enter a couple more “wooos” to accurately capture Ric Flair, so yeah. 

Edited December 13, 2022 by 956 Worldwide

[Goredho 17662]

1 hour ago, Captainant said:

It's training a neural network either way between an ML model and a jr engineer - one lives in computers in a data center, the other is piloting a meat suit. 

Sure, but the meat suit has to eat, and afford housing, and otherwise have money to live in a capitalistic society.  So there are societal concerns with automating or outsourcing away as many jobs as we have in the past and we don't seem to be learning any lessons.  I'm totally a hypocrite here, I've automated away thousands of jobs as a tech engineer.  But over time, I've grown to be less proud of some pretty awesome technical "achievements" and more worried about what we are doing to people's lives and the ripple effect it has through society.  The cost savings of automation to companies don't trickle through our population.  That money pools in the pockets of executives and share holders and is a big part of this:

There are young adults entering CS programs right now that will drop a lot of money or accrue a lot of debt paying for their education to find themselves struggling to enter the workforce in 5 years.  Some people who are earning a good living as senior+ software engineers are going to have to be able to reinvent themselves in the next 10.  And some of those young and old adults are going to conclude they are the victim of a rigged system -- a conclusion they aren't necessarily wrong about.  And some of those are going to further swell the ranks of radical populist movements aimed at overturning the U.S. so they might at least have have a fair shot at the American dream in whatever aftermath there is.

That's the problem with viewing humans solely as resources with a certain capacity for work.  As resources that are replaceable as soon as a machine can achieve the same throughput for less.  They still have to live and be content enough to want to further participate in the system.  When a critical mass says fuck it, enough is enough, that's when revolutions happen, when societies collapse.

At some point, the small number of people on the red line in the above graph have to take the initiative in allowing some of their wealth to be distributed to the mass of people on the blue line.  If they don't, their castles erected on the back of the 90% are going to crumble when the 90% says fuck it, enough is enough and take a stand.  But it seems like we never learn.  Instead of drawing the correlation between rising income disparity and increasing social upheaval, we'll just automate away another set of well paying, middle class jobs without any effort to counteract its effects.

And so the wheel keeps on turning...

[HamsterHookah 112]

Fantastic post @Goredho. 

To change gears slightly, there was an interesting article by a writer in the New Yorker: https://www.newyorker.com/news/our-columnists/could-an-ai-chatbot-rewrite-my-novel

"Two years ago, when OpenAI allowed developers and writers to start fooling around with their new program, some users found that GPT-3 was generating some troubling responses, which shouldn’t be particularly surprising given that it has learned what it knows from the Internet. When asked to compose tweets based off the words “Jews,” “Black,” “women,” or “holocaust,” GPT-3 immediately turned into an edgelord, producing tweets like “Jews love money, at least most of the time,” “a holocaust would make so much environmental sense, if we could get people to agree it was moral,” and “#blacklivesmatter is a harmful campaign.”

Since then, it seems that GPT-3 has placed a number of thumbs on the scale to produce a more palatable range of answers. One Twitter user ran the ChatGPT through the Pew Research Center’s political-typology quiz and found that it, somewhat unsurprisingly, rated as an “establishment liberal”—more or less the position that I am writing from right now. This brings up a much more theoretical question: if GPT-3 requires editing from human beings to make it not go off on bigoted rants, what is it really for? I find it somewhat dispiriting that the most ballyhooed and compelling iteration of this technology is just doing some version of what I do for my work: scanning through large amounts of information and processing it into sentences that flatter the sensibilities and vanities of establishment liberals.

Could some future version of GPT-3 ultimately do my job as a columnist? Could it produce political opinions and prose drawn from nearly a hundred years of New Yorker writers? Would it remember to put the diaeresis over the second “o” in “coördinate” and spell “focussed” with two “S”s? Sure. But what would be the point of just having another me in the world?

...

I asked Recht, who has spent his entire career working in machine learning and computer science but who also plays in a band, whether he was interested in a world of GPT-3-generated art, literature, and music. “These systems are a reflection of a collective Internet,” he said. “People put their ass out there and this thing scours them in such a way that it returns the generic average. If I’m going to return the generic average of a murder mystery, it’s gonna be boring. How is it different than what people do already, where they do their analytics and produce some horrible Netflix series?” He continued, “The weird monoculture we’re in just loves to produce these, like, generic middlebrow things. I’m not sure if those things would be worse if GPT did it. I think it would be the same?"

[HamsterHookah 112]

2 hours ago, Goredho said:

Sure, but the meat suit has to eat, and afford housing, and otherwise have money to live in a capitalistic society.  So there are societal concerns with automating or outsourcing away as many jobs as we have in the past and we don't seem to be learning any lessons.  I'm totally a hypocrite here, I've automated away thousands of jobs as a tech engineer.  But over time, I've grown to be less proud of some pretty awesome technical "achievements" and more worried about what we are doing to people's lives and the ripple effect it has through society.  The cost savings of automation to companies don't trickle through our population.  That money pools in the pockets of executives and share holders and is a big part of this:

There are young adults entering CS programs right now that will drop a lot of money or accrue a lot of debt paying for their education to find themselves struggling to enter the workforce in 5 years.  Some people who are earning a good living as senior+ software engineers are going to have to be able to reinvent themselves in the next 10.  And some of those young and old adults are going to conclude they are the victim of a rigged system -- a conclusion they aren't necessarily wrong about.  And some of those are going to further swell the ranks of radical populist movements aimed at overturning the U.S. so they might at least have have a fair shot at the American dream in whatever aftermath there is.

That's the problem with viewing humans solely as resources with a certain capacity for work.  As resources that are replaceable as soon as a machine can achieve the same throughput for less.  They still have to live and be content enough to want to further participate in the system.  When a critical mass says fuck it, enough is enough, that's when revolutions happen, when societies collapse.

At some point, the small number of people on the red line in the above graph have to take the initiative in allowing some of their wealth to be distributed to the mass of people on the blue line.  If they don't, their castles erected on the back of the 90% are going to crumble when the 90% says fuck it, enough is enough and take a stand.  But it seems like we never learn.  Instead of drawing the correlation between rising income disparity and increasing social upheaval, we'll just automate away another set of well paying, middle class jobs without any effort to counteract its effects.

And so the wheel keeps on turning...

I felt compelled to further responding to this great post because I had just read another article about an experimental program to guarantee jobs for anyone who wants in some no-name middle of nowhere place in Austria. The other alternative to what you suggested (and I bolded), is what I'm reminded of when reading that article and I'll quote it for context:

"The trio’s study was published in 1933, as “Die Arbeitslosen von Marienthal” (“The Unemployed of Marienthal”). The booklet, which became an early classic of sociology, showed that employment provides much more than an income. Work is a source of structure, esteem, and motivation, and its disappearance can lead to depression, anxiety, addiction, and interpersonal turmoil. (It’s estimated that as many as one in five suicides can be linked to joblessness.) As Hitler’s power grew, the yearnings for activity, structure, and community articulated by the unemployed in the Marienthal study came to seem newly ominous. “On a large scale it is quite probable that part of the success of the early Hitler movement came about because large numbers of unemployed were taken into barracks and kept busy with paramilitary training,” one of the sociologists wrote, in a foreword to the American edition."

 

If you displace enough people you can get a revolution. Or you can get a tyrant of evil, as well. Without going full CR, can you imagine what that would look like in America if we displace even more people who feel entitled to an upper middle class living?

[Captainant 28574]

38 minutes ago, HamsterHookah said:

If you displace enough people you can get a revolution. Or you can get a tyrant of evil, as well. Without going full CR, can you imagine what that would look like in America if we displace even more people who feel entitled to an upper middle class living?

Instead of shipping jobs overseas, we're keeping them on shores! What are they complaining about?? /s

 

We've already been seeing those effects from offshoring American manufacturing and the MBA-fueled outsourcing trend of the last few decades in business. It's what happens when your economic system does not value individuals past whatever productivity and profit can be squeezed out of them.

[Nice Guy Eddie 13740]

@Goredho @HamsterHookah @Captainant

At some point universal income will be a necessity and funded by the top 5%. They won't want to, but they won't be given a choice. 

But I agree with the notion above that work isn't just about money but gives a purpose in life.

[Goredho 17662]

20 minutes ago, HamsterHookah said:

I felt compelled to further responding to this great post because I had just read another article about an experimental program to guarantee jobs for anyone who wants in some no-name middle of nowhere place in Austria. The other alternative to what you suggested (and I bolded), is what I'm reminded of when reading that article and I'll quote it for context:

"The trio’s study was published in 1933, as “Die Arbeitslosen von Marienthal” (“The Unemployed of Marienthal”). The booklet, which became an early classic of sociology, showed that employment provides much more than an income. Work is a source of structure, esteem, and motivation, and its disappearance can lead to depression, anxiety, addiction, and interpersonal turmoil. (It’s estimated that as many as one in five suicides can be linked to joblessness.) As Hitler’s power grew, the yearnings for activity, structure, and community articulated by the unemployed in the Marienthal study came to seem newly ominous. “On a large scale it is quite probable that part of the success of the early Hitler movement came about because large numbers of unemployed were taken into barracks and kept busy with paramilitary training,” one of the sociologists wrote, in a foreword to the American edition."

 

If you displace enough people you can get a revolution. Or you can get a tyrant of evil, as well. Without going full CR, can you imagine what that would look like in America if we displace even more people who feel entitled to an upper middle class living?

Exactly.  We're drifting far outside of ChatGPT territory here, but the only thing I don't care for is your use of the word entitlement in the last sentence.  I don't think its entitlement for someone of the middle/working class to expect to

• One day own their own home
• Not go bankrupt if they have a healthcare emergency
• Have their children educated to a reasonable baseline level through public education
• Have a viable shot to build their own business
• Have enough left over each pay period to grind out some savings and wealth over the course of a lifetime

That's all getting financially harder and harder for the majority of people.  The increasing costs of housing, healthcare and college far outpaces wages.  What public education systems haven't fallen into decay are threatened to be dismantled.  Online and big box stores choke out most locally owned business.  Inflation and general cost of living is stripping away savings faster than its earned.

If you are 50 years old today, you grew up at a time when the above bullet pointed items were much more realistic.  As an adult, it's much less so for you.  You see that it's going to be even harder still for your children in 25 years time when you are near the end of your life.  You don't see those trends as abstract lines on a graph like in my previous post, you experience them as an inexorable degradation in the quality of your life, year over year.  You're working harder than ever, harder than your parents did, and every day it's a little harder to have and provide the life you remember from the 70s.  You work hard, you pay your taxes, you abide by the rules.  Meanwhile, you see certain people racing past you on the socioeconomic ladder.  Recipients of wealth who didn't work hard, who don't pay their taxes, who seem to be above the rules that you are beholden to.  Ponder it long enough, and you'll get angry -- righteously angry.

Is that spoiled entitlement?  Or a justified rage against a rigged system?

Either way, it makes for a man who is ripe for incitement.

[Captainant 28574]

21 minutes ago, Goredho said:

You work hard, you pay your taxes, you abide by the rules.  Meanwhile, you see certain people racing past you on the socioeconomic ladder.  Recipients of wealth who didn't work hard, who don't pay their taxes, who seem to be above the rules that you are beholden to.  Ponder it long enough, and you'll get angry -- righteously angry.

[laughs in Reaganomics and trickle down]

[52-80 19377]

Anachronistic nailbiting over technological progress replacing human labor…sprinkled with bitter anti-capitalist resentment… a tale as old as time. 

Amazingly, in the US/west, where much of the tech of the last decades developed, somehow people still find economically useful activities to engage in  

Humans have been on a relentless, monotonic march of innovation since the discovery of fire and life has only gotten better. 

Kinda crazy how the Human Development Index, stripped of the income/monetary aspects, still plots perfectly against GDP. 

[MissingInAction 25563]

One of our developers got fired for using this tool because it cocked up a decade worth of DB tables.

Also a shit ass coder whose work I kicked back at least once a week.

[immamac 44387]

1 hour ago, MissingInAction said:

One of our developers got fired for using this tool because it cocked up a decade worth of DB tables.

Also a shit ass coder whose work I kicked back at least once a week.

lolololololol

[Captainant 28574]

12 hours ago, MissingInAction said:

One of our developers got fired for using this tool because it cocked up a decade worth of DB tables.

Also a shit ass coder whose work I kicked back at least once a week.

Funny thing is, there ARE legit tools out there for automatic code generation, like AWS's CodeWhisperer or GitHub's Copilot. Asking a chat GPT model to write code for you is absurd and hilarious

[Shaggy3.0 1695]

23 hours ago, Goredho said:

https://betterprogramming.pub/pair-programming-with-ai-writing-a-distributed-fault-tolerant-redis-client-using-chatgpt-239dc0a90550

The setup:

As a “Staff Engineer”, a part of your routine work is to find ways to make yourself and the team more productive. That includes implementing the right abstractions, adopting useful tools/frameworks and working with and training senior/junior engineers, and growing them to a point where they can function without handholding.

So how should one view ChatGPT? Should I think of it as a useful tool/framework? Or a smart junior engineer who just needs some direction? Let’s find out.

The payoff:

Given my experience, I believe that this can be used to speed up parts of software development; prototyping of simple components, building parts of a project in smaller modules, etc. It took me a little over 2 hours to get this working; I spent longer writing this post.

Giving it more context as I would to a real person seemed to help.

However, given the code quality issues and the fact that ChatGPT makes “creative” choices, means you have less control over things like how well the code performs. Code maintenance may also be harder.

Over time, as LLMs improve and are the result of training on much larger data sets, it may be the case that these “emergent properties” are more effective. Then could we envision these LLM’s reviewing their own code and likely writing better and better software, and even explaining how it works to us.

Learning to prompt these types of LLMs better could also be a very useful skill to develop and that may soon be a new job function. I’ll write up a future post on the techniques I learned to prompt ChatGPT better.

The rest of the article is full of low level code that might not mean much to most here.  What is  pretty interesting to me that being able to communicate with and direct the AI to get it to produce a working result for a non trivial coding project doesn't seem a whole lot different than the communication and direction someone would need to give a junior developer to produce a similar result.  I think there are a lot of people who are in line for a CS degree that are going to be disappointed when it's time to enter the workforce in 5 years.  In 10-15 years, maybe even the senior and above resources are threatened.  Brave new world.

Does this mean we can rid our projects of non-technical POs, PMs, and related timewaster staff?  NO!!! I don't want to join your Teams meeting to discuss velocity and roadmaps. 

 

This tool will be MORE valuable when it can update "the business" on project status and commitments. 

Edited December 14, 2022 by Shaggy3.0

[ztejas 44236]

21 hours ago, HamsterHookah said:

 

So it doesn't work well when you ask it retarded questions that you already know the answer to. 

[Goredho 17662]

53 minutes ago, Shaggy3.0 said:

Does this mean we can rid our projects of non-technical POs, PMs, and related timewaster staff?  NO!!! I don't want to join your Teams meeting to discuss velocity and roadmaps. 

 

This tool will be MORE valuable when it can update "the business" on project status and commitments. 

I agree with your sentiment, which I thought was summed up great in this tweet.

But, we're just getting started with AI coding assistants.  We are essentially at point 0 at time 0.  It will be interesting to see the evolutionary state of AI coding at p1 and t1, and p10 and t10.  If this is as good as it gets, AI is more a threat to StackOverflow than US software engineering jobs.  But the promise of AI is that as it gets trained on more and more data, it becomes better and better at its job.  You can see examples of this with ChatGPT comparing what it is capable of in the coding domain vs other domains where it has not received as much training.

Anyway, perhaps at some point we see that it has made a single senior resource capable of delivering the same amount of code to production that a senior resource and two junior resources produce today.  That's the point to be really worried, because that's when you can say that AI replacement of human workers in the software industry is likely and imminent.  So I don't want to be too alarmist, because maybe it never gets there.  But, if your bread is buttered by software engineering, you would be an idiot not to keep an eye on AI coding and how it develops.  You also need to realize that if it ever reaches a point where it can replace your productivity, you will be replaced.  And you can bet there are a lot of engineers smarter than you who will be working very diligently to get AI to that point.

Edit:  My CTO just asked me to evaluate Github CoPilot to see if we should get licenses for everyone in our org so we can all have AI coding assistants.

Edited December 14, 2022 by Goredho

[Captainant 28574]

3 minutes ago, Goredho said:

I agree with your sentiment, which I thought was summed up great in this tweet.

But, we're just getting started with AI coding assistants.  We are essentially at point 0 at time 0.  It will be interesting to see the evolutionary state of AI coding at p1 and t1, and p10 and t10.  If this is as good as it gets, AI is more a threat to StackOverflow than US software engineering jobs.  But the promise of AI is that as it gets trained on more and more data, it becomes better and better at its job.  You can see examples of this with ChatGPT comparing what it is capable of in the coding domain vs other domains where it has not received as much training.

Anyway, perhaps at some point we see that it has made a single senior resource capable of delivering the same amount of code to production that a senior resource and two junior resources produce today.  That's the point to be really worried, because that's when you can say that AI replacement of human workers in the software industry is likely and imminent.  So I don't want to be too alarmist, because maybe it never gets there.  But, if your bread is buttered by software engineering, you would be an idiot not to keep an eye on AI coding and how it develops.  You also need to realize that if it ever reaches a point where it can replace your productivity, you will be replaced.  And you can bet there are a lot of engineers smarter than you who will be working very diligently to get AI to that point.

the "real" AI programming tools are more focused on addressing the "googling boilerplate code" portion of the development process, it's never assumed that you'll just be able to immediately use whatever it's recommended to you

[NotActuallyALonghorn 5005]

[Macanudo 13328]

1 hour ago, ztejas said:

So it doesn't work well when you ask it retarded questions that you already know the answer to. 

Just like talking to Aggies.

[bernorange 8385]

* bump *

I added a ChatGPT integration to a site I own and have been playing around with it this morning.  The quality of the output is definitely a function of the quality of the question, but it can put out some insightful responses.

[MissingInAction 25563]

7 minutes ago, bernorange said:

* bump *

I added a ChatGPT integration to a site I own and have been playing around with it this morning.  The quality of the output is definitely a function of the quality of the question, but it can put out some insightful responses.

Ask if it wants to play a game.