vendiddy

I may have used the term loosely.

When I was learning Spanish, the other person I was speaking to would simply repeat what I was trying to say in Spanish.

It wasn't always nitpicking my grammar. In many cases it was predicting what I was struggling to say.

I would have loved to be in your class because it sounds a lot more focused on practical conversation than my first language class!

That sounds interesting. Thanks for the recommendation!

I'm describing how I learned Spanish.

Spanish is not a programming language.

I'm neither, but I tried learning Spanish from a "language teacher" and it didn't work. I'm not alone. I've spoke to dozens who had the same experience.

Years later I learned from people with the same method I described.

The AI models (large LANGUAGE models) of today seem very close to replicating that experience.

Try telling ChatGPT with voice mode "let us talk about X in [pick language]... Correct me when I'm wrong and explain why".

I'm curious where you think it's failing.

What do you feel is missing to replicate a good language teacher?

The programming model of message passing between processes lends itself well to certain types of apps even if you don't need to handle millions of connections.

For example Elixir/Erlang are currently used in certain IoT applications.

How is the security?

Can't someone learn this stuff with the assumption that certain functions are hard to invert?

Like assuming f(x) is hard, let us try to prove these other properties of security.

Nix specifically may have governance problems, but this connection seems like a stretch to me

I do think AI could be a better language tutor than an average language teacher, but I don't think Duolingo's approach is very effective.

The ideal AI-powered tutor would work more closely to a private language tutor. It would speak with you and gradually integrating language concepts into the conversation. When you make mistakes, it could correct you on the spot and keep track of where your strengths and weaknesses are.

You're not the market

I don't find this to be absolutely true.

At our company we reinvented a handful of wheels because the existing solutions were too complex.

And complexity slows you down.

It's not this clear cut.

I really like doing this.

I find that in many cases you can cut out 80 percent of the code of the original library.

Most of the deleted code is flexibility and features we don't need.

It's surprising how small the essence of a solution can be.

I've used it these past few months to better understand the PDF format, Nix, and a few other technical concepts.

I try to use AI to automate things I already know and force myself to learn things I don't know.

It takes discipline/curiosity but it can be a net positive.

How do you find Aider compares to Claude code?

I think they are capable of doing it, but it requires prompting.

I constantly have to instruct them: - Go step by step, don't skip ahead until we're done with a step - Don't make assumptions, if you're unsure ask questions to clarify

And they mostly do this.

But this needs to be default behavior!

I'm surprised that, unless prompted, LLMs never seem to ask follow-up questions as a smart coworker might.

Managing complexity pays off sooner than one would think.

Even a project that's like 15k lines of code would benefit from a conscious effort to fight against complexity.

Yep I want Zed to win but it has not yet become my daily driver

What makes this tough is there are so many variables.

What an AI could not do last week might be possible this week.

Or maybe my prompt didn't have enough guidance.

For the time being we have to keep testing the waters.

That's a good one.

One analogy I use a lot: if I have a professor sitting next to me, what is the best way to learn a topic?

Struggle through it on my own and I won't be leveraging the professors knowledge.

Ask the professor to do everything for me and I won't be learning anything at all.

Now if the professor is an AI, the same trade-offs hold.

For example, I will back and forth conversations with AI to explain subjects to me. I ask questions, push back, ask for examples, and so on.

If I do ask the AI to answer something for me, I then ask it to break down the answer for me so I can make sure I understand it deeply.

And of course, none of this matters if I don't want to learn something :)

No worries I did not take it that way!

I may not have been clear, but it was given to us as a definition in class. As in we saw the matrix and the corresponding equation.

But nobody even told us why! And at that time I never thought to ask.

It took me a long time to realize math consisted of abstractions.

I was never told, for example, that matrices are a useful abstraction (shorthand?) for representing linear equations.

Or that imaginary numbers were an invented abstraction that made certain calculations easier.

It would have been nice to learn it from the perspective of how those abstractions came into being and why they're useful.

As a lay person, apps like Office feel like they stay the same between updates and somehow manage to get fatter and slower.

(I'm sure they've got some new improvements tucked away but I don't notice them.)

Or every time I update my OS it's like a 10GB download. What did I just get for that 10GB? I honestly can't tell.

Even my smartphone. Seem often where I'm asked to install a 2GB update aaand it's same as before but slower?

Anyone else have this sentiment?

I love coding but I also love AI.

I don't know if I'm a minority but I'd like to think there are a lot of folks like me out there.

You can compare it to someone who is writing assembly code and now they've been introduced to C. They were happy writing assembly but now they're thrilled they can write things more quickly.

Sure, AI could lead us to write buggier code. Sure, AI could make us dumber because we just have AI write things we don't understand. But neither has to be the case.

With better tools, we'll be able to do more ambitious things.

Those oauth permission dialogs come to mind but they may need to improve significantly to become useful.

I don't think most of those are granular enough.

It's easy to forget how tedious things used to be before React became popular.

Keeping data in sync with the UI was a huge mental burden even with relatively simple UIs. I have no desire to go back to that.

I think this is spot on.

I feel like folks on both sides would stop talking past each other if they were willing to understand the other POV rather than dismissing it as crazy.

I think AI will have a dual effect. It will make some folks smarter and others dumber.

For example, you could have ChatGPT write your code for you, then explain it to you step by step.

It can be an interactive conversation.

Or you could copy/paste it.

In one case it acts as a tutor.

In another case it just does your work for you.

I'm very happy that something like this exists!

I'm wondering why have there have been no good IDE experiences so far for Postgres? Or put another way, what has been the most challenging part of building this?

Nothing I have tried so far comes close to what I'm used to with statically typed languages. One would think something as strict as Postgres would have good autocomplete by now but I've yet to find something.

I'm glad I'm not the only person in awe with LLMs. It feels like it came straight out of science fiction novel. What does it take to impress people nowadays?

I feel like if teleportation was invented tomorrow, people would complain that it can't transport large objects so it's useless.