Is programming a worthwhile skill to learn for entrepreneurship?

[Gubolo]

Everyone knows AI is gonna be a big thing. I've heard a lot of people say that AI is going to replace programmers. What do you guys think? Is programming still going to be a good skill to learn. Do you have any ideas for how to prepare yourself to use AI so you can start a AI-based business?

 

Dislike ads? Remove them and support the forum: Subscribe to Fastlane Insiders.

[heavy_industry]

I wonder how long will this mass hysteria last.

 

[BellaPippin]

 

[Rangermac2]

Everyone knows AI is gonna be a big thing. I've heard a lot of people say that AI is going to replace programmers. What do you guys think? Is programming still going to be a good skill to learn. Do you have any ideas for how to prepare yourself to use AI so you can start a AI-based business?

AI Currently uses scrapers to collect data and then makes assumptions to piece it together. It does not entirely think and make programs like you and I. We can discover solutions but AI reveals them since it has a large body of knowledge. I've used ChatGPT to help with some python indexing and data frames, which was really helpful. But it's not at the level that it can make extremely accurate assumptions based just on intuition. It relies on the data that already exists and then uses Neural Networks and Decision Trees to make an assumption.

Machine Learning has been around since the 1950's with conditional statements. Now there's also Deep Learning which uses data and neural networks which was made in the 1940's by Walter Pitts and Warren McCulloch. So really AI isn't anything new. It's just the implementation and sophistication that is.

Use AI as a support system and to help find solutions, don't rely on it entirely. It'll be a lot longer before it can perform tasks like programming reliably on it's own, while making its own concepts and solutions. I Hope this helps

 

Dislike ads? Remove them and support the forum: Subscribe to Fastlane Insiders.

[M.B.W]

I own a copywriting business and AI absolutely obliterated my company. So, I know a thing or two about this.

Our main focus was high volume SEO optimised copy. It basically shrunk the value of our service by 70% within 3 months, and completely saturated the market.

I believe that programming won't become obsolete as a whole, but experienced programmers who are manning large projects will be using AI instead of hiring a team.

 

[AppleTree]

Being a programmer isn't just being able to code, a lot of the job of a programmer is communicating with clients and designing projects. At the current stage, the AI is still too novice in those departments and can realistically only replace very junior programmers. It's main purpose now is just replacing the more mundane repetitive tasks. Machine learning and deep learning has been around for a long time(Github Copilot was out way earlier than ChatGPT, ChatGPT just recieved this much attention because it was a model speciailized in communications) and although it's making huge improvements, it's definitely not at a level where it can fully replace a working programmer. Honestly if it is able to, then we have more things to fear about than just programmers being replaced.

 

[Absar]

I was trying to solve an error in my code, and asked GPT with 10+ different prompts, every time it showed a solution that doesn't just work or is outdated.

So No, and this is a stupid question.

 

Dislike ads? Remove them and support the forum: Subscribe to Fastlane Insiders.

[Gubolo]

It is just because I just started to try learning programming and I was wondering if I should quit or continue.

 

[AppleTree]

It is just because I just started to try learning programming and I was wondering if I should quit or continue.

It depends on your fastlane path. If you want to create a business that needs coding, learning programming could help you create your prototype. You could also go on the path of just hiring programmars to create the prototype and go on from there.

 

[Two Dog]

Everyone knows AI is gonna be a big thing. I've heard a lot of people say that AI is going to replace programmers. What do you guys think? Is programming still going to be a good skill to learn. Do you have any ideas for how to prepare yourself to use AI so you can start a AI-based business?

You prepare yourself to use AI by becoming a world class expert at using AI. That's a skill that doesn't exist yet.

The reason it seems like AI will kick everyone's a$$ in every imaginable way is that *everything* you've seen to date is pretty simple to make happen. When someone can diddle for ten minutes and get something impressive, it means anyone can diddle for ten minutes and get something impressive. Which quickly makes similar results NOT impressive to anyone.

The real power of AI will start becoming apparent when there's a big enough group of skilled users to doing things that almost no one can imagine right now. When it comes to programming - and any skill for that matter - AI will certainly change what the person does every day. Instead of "replace" them, it's far more likely to "displace" them to do other tasks at a much higher level of performance and productivity, but that requires a solid base of knowledge and expertise in the first place.

Most people will be largely incompetent using AI. That's not because of AI. They're incompetent at most things because they're unwilling to put in the time and effort to become an expert. Non-programmers will not suddenly outcompete skilled programmers because of ChatGPT. Non-lawyers will not suddenly outcompete skilled attorneys. Non-copywriters will not suddenly outcompete professionals.

Funny example...

Friend of mine was playing around with ChatGPT to write sales copy for his website. He was very excited that it could generate a page of decent copy with minimal editing in only a couple of minutes. Great start. However, it took him almost two months to reach that point because he's not particularly motivated and busy with other things. I pointed out that he's published barely any content - much less anything connected to improving the site - over the past three years. That two minute blog post could have been purchased for five bucks at any time.

Finally, I asked him: "If ChatGPT is so goddamn smart, why are you asking it to write a handful of blog posts? Why don't you just ask it to create a fully functioning website that generates $1,000 / month for you?" Dead silence.

 

Dislike ads? Remove them and support the forum: Subscribe to Fastlane Insiders.

[The-J]

I've heard a lot of people say that AI is going to replace programmers.

When you hear a lot of people say something about the future, it's probably stupid and wrong.

 

[OMDA]

I've thought about this some. I still think programming, even if it's assisted by AI, is worth having as a skill.

I also think about something related to my work: mechanical engineering analysis software. It can take days to run a realistic analysis solve on a cluster. And it can still be non-physical and wrong. It takes an expert to spot the errors. AI to run the simulation would be incredibly expensive right now.

BUT, also, there were engineers that with just side rules, drawing skills, and knowledge, that were able to make us a spacefaring species and design quite a bit of the sr-71. Good design of things require judgement. Unfortunately, without real world experience, it is very hard to come by this computer age. The need for good and experienced people to bridge the gap from the digital to the physical is there today and probably not going away for a while.

 

[amp0193]

I've heard a lot of people say that AI is going to replace programmers. What do you guys think?

Eventually, yes.

Is programming still going to be a good skill to learn.

Yes, because until it does, there's lots of opportunity left to take advantage of. Especially in niche solutions to problems that others won't bother to solve.

It will take a while yet for AI to reach into every nook and cranny.

Do you have any ideas for how to prepare yourself to use AI so you can start a AI-based business?

Train hyper-niche LLMs designed for a specific purpose. Look at what people are doing with LLaMA and open source.

Also middle-layer that uses ChatGPT API to tailor it to specific function or SAAS.

No advice, other than to learn all you can, as fast as you can. And look for a need that needs solving.

 

Dislike ads? Remove them and support the forum: Subscribe to Fastlane Insiders.

[Saad Khan]

Train hyper-niche LLMs designed for a specific purpose. Look at what people are doing with LLaMA and open source.

Also middle-layer that uses ChatGPT API to tailor it to specific function or SAAS.

Man, you really crystallized my idea and put it into words! I'm learning AI so I can train and tweak these LLM's for a specific niche. Like an AttorneyGPT for US people looking to consult with a lawyer.

And to answer OP's question...

No. It won't replace programmers. But a programmer who's an expert in using AI will replace you.

 

[amp0193]

Like an AttorneyGPT for US people looking to consult with a lawyer.

yeah you've got the right idea, but you gotta get 100x more niche and specific than that.

And you've got to get your hands on some quality niche data sets... the more gated / hard to get / specialized / technical the better.

There's an opportunity there too for the dealmakers here... AI data set sales broker. Bring buyers/sellers together, broker the deal, keep the commission.

No. It won't replace programmers.

Until AGI

 

[Saad Khan]

yeah you've got the right idea, but you gotta get 100x more niche and specific than that.

100%

And you've got to get your hands on some quality niche data sets... the more gated / hard to get / specialized / technical the better.

That will be the secret sauce for the supervised learning models.

There's an opportunity there too for the dealmakers here... AI data set sales broker. Bring buyers/sellers together, broker the deal, keep the commission.

This is a big opportunity for people who aren't AI researchers/engineers but keep a pulse on the market.

 

Dislike ads? Remove them and support the forum: Subscribe to Fastlane Insiders.

[Gubolo]

But doesn’t programming take a while to learn? Is it like a thing you should go all in and dedicate your entire day to learning, or like a thing you learn while also taking action with something else?

Man, you really crystallized my idea and put it into words! I'm learning AI so I can train and tweak these LLM's for a specific niche. Like an AttorneyGPT for US people looking to consult with a lawyer.

And to answer OP's question...

No. It won't replace programmers. But a programmer who's an expert in using AI will replace you.

I'm trying to adapt my programming learning to AI. I've asked my programming "coach" how I can adapt it to AI.

 

[theplannerr]

It would do so when AGI (if it does) becomes available;
As Angel Investor Naval says, "AI is a tool - it replaces repetitive work, not creative work."

Only programmers who would be replaced would be those lowest at the totem pole.

 

[Gubolo]

Eventually, yes.


Yes, because until it does, there's lots of opportunity left to take advantage of. Especially in niche solutions to problems that others won't bother to solve.

It will take a while yet for AI to reach into every nook and cranny.



Train hyper-niche LLMs designed for a specific purpose. Look at what people are doing with LLaMA and open source.

Also middle-layer that uses ChatGPT API to tailor it to specific function or SAAS.

No advice, other than to learn all you can, as fast as you can. And look for a need that needs solving.

I don't really understand what any of that last stuff means, but I will try to get as good with AI as fast as possible.

 

Dislike ads? Remove them and support the forum: Subscribe to Fastlane Insiders.

[amp0193]

But doesn’t programming take a while to learn? Is it like a thing you should go all in and dedicate your entire day to learning, or like a thing you learn while also taking action with something else?

Takes as long as mastering any other skill.

10,000 hours

Find the time to do it or don’t.

don't really understand what any of that last stuff means, but I will try to get as good with AI as fast as possible.

Google and read.

You gotta be knee deep in this stuff if you intend to find opportunity to add value.

As Angel Investor Naval says, "AI is a tool - it replaces repetitive work, not creative work."

For now.

What is creativity though? What makes humans uniquely capable of doing it? Is there some secret sauce to the human brain that can’t be replicated with enough computational power and data?

Extreme pattern-matching and synthesis.

I think there will be a point in which AI will be a lot more creative than us, because our hardware is limited to a slow-processing organic substrate and a single life’s worth of input data.

 

[Matt Sun]

New tools bring new opportunities for problem solving:

This agricultural company enables organic chemical free farming by using AI to recognize weeds and blast them with a laser, this would replace regular weed killers like roundup. Says it can kill 100k per hour. Pretty amazing.

I'm pretty sure they needed / will need programers to run the company.


View: https://www.youtube.com/watch?v=vSPhhw-2ShI

 

[Two Dog]

Until AGI

AGI = Artificial General Intelligence.

It's so generic I didn't expect Google to provide a useful answer, but it found Sam Altman's quote easily enough.

A blog post from CEO Sam Altman describes AGI as anything “generally smarter than humans.”

Y'know, it really seems the Silicon Valley tech bros are really super smart about solving immediate problems and dumber than rocks when it comes to understanding their implications. WTF does "smarter" even mean when everyone has a different takes on what they'd like to see happen.

I'm trying to adapt my programming learning to AI. I've asked my programming "coach" how I can adapt it to AI.

I'm reluctantly getting around to modernizing a creaky old SaaS app largely written with techniques that are about 10-15 years old. One of the biggest reasons for putting it off was the sheer amount of work involved much less finding someone able to do the work. Rewriting from scratch is almost always a mistake. I've vaguely paid attention to development trends, but stopped somewhere in the world of Laravel, Vue and Node thinking that's the best approach nowadays.

After only a few hours of research, it's pretty clear that if you're learning to code without using the tools developed over the past 12 - 24 months, it's a complete waste of time. No code, low code, pre-built templates, Github Copilot, all the tools that let someone put up a fully functioning mobile app, web app, whatever and have it scale to infinite users are the ONLY thing to spend time learning. Unless you're planning a career in modernizing existing web apps (and there's a GIANT market worth considering), no one should learn to code the way I did.

 

Dislike ads? Remove them and support the forum: Subscribe to Fastlane Insiders.

[Beerbread]

I think it helps if you're already an expert in your field, because AI is definitely not ready to replace certain industries. Even as a bookkeeper, I still have to occasionally correct ChatGPT on how to code certain accounts and some journal entries.

I see it as a good compliment and a good tool to the right direction, but not the end all.

 

[Gubolo]

AGI = Artificial General Intelligence.

It's so generic I didn't expect Google to provide a useful answer, but it found Sam Altman's quote easily enough.

Y'know, it really seems the Silicon Valley tech bros are really super smart about solving immediate problems and dumber than rocks when it comes to understanding their implications. WTF does "smarter" even mean when everyone has a different takes on what they'd like to see happen.


I'm reluctantly getting around to modernizing a creaky old SaaS app largely written with techniques that are about 10-15 years old. One of the biggest reasons for putting it off was the sheer amount of work involved much less finding someone able to do the work. Rewriting from scratch is almost always a mistake. I've vaguely paid attention to development trends, but stopped somewhere in the world of Laravel, Vue and Node thinking that's the best approach nowadays.

After only a few hours of research, it's pretty clear that if you're learning to code without using the tools developed over the past 12 - 24 months, it's a complete waste of time. No code, low code, pre-built templates, Github Copilot, all the tools that let someone put up a fully functioning mobile app, web app, whatever and have it scale to infinite users are the ONLY thing to spend time learning. Unless you're planning a career in modernizing existing web apps (and there's a GIANT market worth considering), no one should learn to code the way I did.

I have a subscription to a webiste called hyperskill, and that is how I'm learning it. My mom has a friend who is a computer scientist and he is helping me learn it when I need it.

 

[Two Dog]

I'm trying to adapt my programming learning to AI. I've asked my programming "coach" how I can adapt it to AI.

I have a subscription to a webiste called hyperskill, and that is how I'm learning it. My mom has a friend who is a computer scientist and he is helping me learn it when I need it.

Yeah, there's a bunch of them nowadays.

The best idea is to build a shareable public project, not just run through a course curriculum.

Learn only what you need for the project, not how to become a Top 1% Javascript Developer.

 

Dislike ads? Remove them and support the forum: Subscribe to Fastlane Insiders.

[WillHurtDontCare]

AI will make everything obsolete - just give up now and start eating the bugs in your pod

 

[amp0193]

Here’s another AI opportunity: ignore it, and the things it can easily replace, and sell physical products instead.

Even as a bookkeeper, I still have to occasionally correct ChatGPT on how to code certain accounts and some journal entries.

Many of my Ecom friends have ditched their bookkeepers completely and use finaloop now. They say it’s super good. Missing some higher touch functionality but a lot of that is in the works. I will probably do the same soon.

 

[Xavier X]

Will AI replace all programmers? Not a chance.

Will AI replace a ton of entry-level programmers? With 99.999% certainty.

ChatGPT is already doing that, in under a year. For 10+ years, I used to hire freelance programmers to code certain scripts and plugins for me. Since ChatGPT, I just do most of it myself now. All the people I've been giving basic programming work for years have suddenly been replaced by AI.

Sure, there's a lot of "mass hysteria" around AI this and AI that. However, many people are making the mistake of judging AI's potential using today's capabilities and shortcomings. It's like being in 1890 and arguing if a passenger carriage will ever reach 30 mph. Many rationalized the physical shortcomings of horses into their assessment, without consideration for engine-driven carriages. Today is not tomorrow.

 

Dislike ads? Remove them and support the forum: Subscribe to Fastlane Insiders.

[amp0193]

Since ChatGPT, I just do most of it myself now.

I don’t need to hire Shopify coders for almost anything anymore.

Anything I can imagine that I want to do in excel, chatGPT gives me the formula.

And chatGPT still kind of sucks.

Imagine the impact it will have in the near future when it sucks a lot less.

 

[EmotionEngine]

I don’t need to hire Shopify coders for almost anything anymore.

Anything I can imagine that I want to do in excel, chatGPT gives me the formula.

And chatGPT still kind of sucks.

Imagine the impact it will have in the near future when it sucks a lot less.

That's exactly what it's good for. I know can program fairly well and it's a great assistant especially when you know what to ask it. Only programmers can unlock it further with prompts and get more out of it because they know a specific language.

Asking it to create a new soda that destroys Coca-Cola or come up with a videogame that sells better than Call of Duty and it will fail every time.