From my experiences, any output I got from "AI", required already some level of experience/understanding of the tooling or language to get something useful out of it.
But there is a middle ground between writing your own code and buying an off the shelf app - no/low code tools. For example, you can use our Easy Data Transform software to drag and drop a data processing pipeline to turn data from one form to another, no need to learn (or ask an AI about) Python+Pandas or R. But this only an option if there is a tool fits your needs, off course.
She couldn't make a business application and go out and start selling subscriptions, but there are whole categories of tools that can be created that are ultra simple CRUD apps that bring specific value to one person. Code quality and/or understanding how it works is irrelevant for that use case.
Until you encounter an error that needs debugging, or you want to extend their functionality.
A child can fill in tons of gaps with their animation (see: Rugrats). We're talking about "is AI good enough to enable a 100% non-technical user able to create software on the fly without the non-technical user losing patience?" This is the discussion.
edit: they would use ChatGPT to ask questions/find direction but wouldn't actually use the code if any was given but I think if you're using say WP where you can just dump in random code via a plugin, people would try that
I just don't see "AI" spitting out code and people just putting it somewhere and it all working just fine, without having to learn some of the basics first.
Maybe some highly integrated system could work, that does the doing for you, and users don't _have_ to deal with code and the tooling to run it. But those systems will again be quite limited, so questionable how "personal" it gets.
One thing I've used "AI" for multiple times already with little to no modification, is to write some one-off script to automate some basic repetitive task (e.g. batch converting something with ffmpeg). Can you call this personal software?
It is crazy what Zapier does. I learned Airtable's API myself and setting up the webhooks took me a bit where as it's just there for Zapier. Started to make me think which way is faster do it myself or use this bridge granted Zapier costs money but yeah.
So funny how things go we wanted to use Airtable now they're like "we want our own branded Airtable" and eventually we'll probably just go with a hosted DB like RDS anyway ahh well.
> I pasted the code you gave me into my Wordpress and now when I refresh the page it’s just white. Why?
edit: I've seen this as a potential answer before
I once heard an anecdote that being a welder makes you look at metal objects differently than non-welders. What a non-welder sees as rigid and inflexible a welder sees as able to be changed.
I think the view of software for programmers vs. non-programmers is similar.
Having said that, just as most people don't see the benefit to being a welder, most people aren't going to suddenly see the benefit of being a programmer.
Was it this one?
“One of the unexpected things about watching the steel guys work is how the solidity of metal means nothing to them. Most people think of metal as something hard and inflexible, but welders don't. Which should be obvious in hindsight, I guess. But, for example, they have these saw-horses that are made of tube steel. And I can see how that came about: they needed some saw-horses; they had some steel. It took them 30 seconds to make them. And, an example with the stairs: the legs of the stairs' landing platform have big threaded bolts for feet, to fine-tune the height of the legs for levelling. And there are these steel tube sleeves that go around the legs, that drop down and cover the bolts. So when they were moving this platform in, they had to flip it over, and they didn't want the sleeves to fall off while they did this. Now to me, that job calls for duct tape. To them: they welded the sleeves in place, then de-welded them when they needed them to move again.”
Absolutely. I tried to search-engine it when I posted and couldn't come up with it. Your search-engine skills are superior!
I think this is true, but it also actually requires you to want to make your own software. In cooking, most people are aware that they're physically capable of making a roast chicken with gravy and mashed potatoes.
Yet takeaway/delivery is thriving. It's not that the ability to cook is scarce, but the desire to do it. I think the same is (now) true with building software.
A cookbook lets you cook like Thomas Keller or Eric Ripert or whoever because you get their exact recipes, even when those recipes have 18 steps and take 7 hours.
An air fryer lets anyone make something acceptably tasty in 20 minutes and with zero technique.
Imo this is what AI is doing for software building. It lets almost anyone build something that accomplishes a job. But we don't have to expect it to have scalable architecture, beautiful UI or follow security best practices.
I think part of the problem is that most software engineers rightfully care about these things for their jobs. But maybe they don't matter if you're building "home-cooked software".
The same way you can read cookbooks from 3 michelin star chefs and wonder if anyone will ever follow the recipes in them (check out Marc Pierre White recipes for an example).
Cooking (manually) takes skill (let's say that's a one-time thing you need to have acquired in the past) and then it needs time (repeatedly).
The outcome (food) is a temporary benefit.
Software meanwhile, in general, is written once in order to solve a problem at all, or make a task easier/more repeatable. The software is the cook, so to speak.
A more apt comparison would be to craft a cooking tool that makes your work easier, or possible at all. (e.g a pot)
We've already had solutions to this for years, it's called jamming a bunch of plugins on to a Wordpress site so it roughly does what the person wants it to do.
or perhaps remember Visual Basic?
Why does every god damn thing on the website need to be about how "AI will change everything"? You do realize we've had these tools since the 90s, right?
There's a reason only a small percentage of people are actually interested in making their own software. Most people just want to use the thing and be done with it, not tinker around with code.
How many people are making their own clothing? Clearly sewing machines are widely available, so they will surely create a revolution where everyone will make their own clothing! And yet - do you want to actually make your own clothing? No. You want to have time for other things, like making your own software.
Yes, but it is a way. And it's a way which enables many more people to start going it, then other ways. Maybe they will also never leave this way, or reach great heights on it, but that's ok, if they get their things done.
> How many people are making their own clothing? Clearly sewing machines are widely available, so they will surely create a revolution where everyone will make their own clothing! And yet - do you want to actually make your own clothing? No. You want to have time for other things, like making your own software.
I can buy cheap clothes in any shop. Usually cheaper and in better quality than making it myself. But AI is more or less for free, there is no investment. And It's useful for gaining solutions which one other would not get.
The more appropriate comparison here would be probably Excel, which is also used for personalized "software" by people.
Eh... While I agree in general, let's not compare dragging controls on forms and writing sometimes pretty arcane code incantations to writing a simple request in plain English.
Does the average person even use computers outside of desk workers at work now, anyway? Can you build and deploy an app with an iPhone/Android app?
It's not even calibrated to speed or anything I just have an ADC connected to a DC motor and logging the values/mapped to milliwatts via resistor value
I also had a solar cell but the coolest thing with that is seeing that bell curve or solar maximum
On the topic of personal software, I keep making note taking apps. I've made chrome extensions, mobile apps, Android widget, desktop app, etc... with multiple databases, end of the day it's just grabbing text, though recently I did go advanced and add in drag-drop image to base64 support (lazy) at some point I'm trying to unify all this data into one place with a sync mechanism
The reason we even need AI to build "personal software" is because building modern software is terrible. Building for the web is terrible.
If we had a modern OS with easy to use app building tools and a language as easy to learn as smalltalk, then anyone could make software, including children. You can add AI on top of that and it's still better because at least you can understand what the AI made.
I posted about it recently: https://news.ycombinator.com/from?site=cgarduno1garduno.gith...
Here's the repo: https://github.com/cgarduno1garduno/SPOT
The thing that blew me away was that this took ~2 hours total time and I didn't touch the code at all - this was 100% done using Cursor. Similarly, I've been building other tools that I use at work and my team is starting to roll them into our standard workflows because they work well.
There are many instances where I struggle to get it to make what I want, but overall I've been really impressed by how much it has helped me and by how easy it has been to start building my own personal software.
I just don’t buy the supposition that in ten years those incentives will still attract the common man.
Love this take. After over a decade of software development I've gained much more appreciation for well-built/useful products.
Yes build personal software. But to act like someone can leverage AI to do all the donkey work without actually understanding it is just silly.
Yes, an app can be a home-cooked meal. Back in the day, we didn't feel the need to eat out all the time.
It gave you the freedom and reduced the friction of iteration to figure out how you needed to organize things to discover what you wanted.
But it brought many limitations that surprisingly (to me at least) still persist today. I think AI has the ability to give you the same kind of freedom and friction reduction to allow you to build what you need.
Yeah, because that wouldn't be a customer service / troubleshooting nightmare.
> But within the next decade, millions of people will be able to create software. Designers, marketers, product mangers, and others will be the first.
Hahaha - clearly the author is young. All of this has happened before, and it will all happen again.
…yet the author's points are completely invalidated by the lack of any explanation for how they intend to address the myriad of horrifying ethical lapses involved in how these models are developed and deployed—not to mention why most of the key influencers in this "AI" movement are deeply anti-craft and anti-humanist.
We in the pro-craft protest movement are happy to engage in good-faith conversations about which types of "AI" technologies might be good at in the future given satisfying resolutions to the host of ethical issues currently involved, but articles like this which seem shockingly unaware of this fierce debate simply come across as uninformed, and uninformative.
I'm so tired of the relentless hype around AI. No normal non-technical person is going to build and be able to maintain their own software.
It's very easy to see where a lot of VC money is going: marketing.
