itsjoshlee.bsky.social - Profile | ThreadSky | a Reddit-style client for Bluesky

Are you getting into bed with AI too quickly? It's a tale as old as time... A business adapts a new technology before fully understanding it... While the experts who do understand it try to pump the breaks a bit to make educated changes.

submitted 40 days ago • 1 comment

At the end of the day, we're building software for people (most of the time anyway...) and if you understand people well, you'll write better software.

submitted 40 days ago • 0 comments

Product market fit is more important than scalable code when starting out.

submitted 41 days ago • 0 comments

A core part of generative AI is that the output is based on what's most likely according to its dataset. Its output is essentially "give me the most probable output according to my input". Add that to users who aren't creative and give minimal input to the LLM...

submitted 41 days ago • 1 comment

I never worry about accidentally writing bad code... I just make sure all my code is bad so I'm not surprised by anything.

submitted 42 days ago • 0 comments

Get market feedback before spending too much time building your product so you know you're building a product users actually want.

submitted 42 days ago • 0 comments

The job of a software engineer is to solve problems, not to write code. Writing code is just usually the way we solve those problems. It's similar to a carpenter - does a carpenter spend all day thinking about how to hammer nails or do they focus on building houses?

submitted 43 days ago • 0 comments

Too many startups worry about setting up their tech to scale with 10,000 users when they don't even have 10. Build, get market feedback, generate revenue, grow a user base, and then worry about how to scale when you actually start scaling.

submitted 43 days ago • 0 comments

There's a huge opportunity in AI right now nobody is talking about... An opportunity that anyone can take advantage of.... and it's much easier than you think. Generative AI is good at a lot of stuff, but there is one thing AI can't do... And if YOU do it, you're going to stand out...

submitted 44 days ago • 1 comment

I would hate to know the new software my bank is using was "vibe coded."

submitted 44 days ago • 0 comments

Going into a legacy codebase and instantly making huge changes because you think the code is bad or that you know better usually isn't a good idea. There's probably a reason the code was written that way and the authors of the code have probably considered your amazing solution before, but didn't

submitted 45 days ago • 0 comments

I like to focus more on 'shipping products' than 'writing code' even though I need to write code to make the products. I think it's a real mindset shift that a lot off developers don't consider.

submitted 45 days ago • 0 comments

As engineers, we want to write - and work with - good code. Code that's easy to understand and use. We're masters at our craft and don't want to push bad code. After all, sooner or later, that technical debt is gonna come back and bite you. But that's not the whole picture...

submitted 46 days ago • 1 comment

When they're a 10 out of 10, but write their functions like this: function foo(bar) { }

submitted 46 days ago • 0 comments

This might make some of you mad, but I need your opinion. Is this an accurate description current AI models? 👇 I go to your house, use your water hose to fill up my swimming pool, charge people to swim in my pool, and then go on podcasts talking about how great my water is

submitted 47 days ago • 0 comments

"It doesn't scale." You don't even have users yet! There's nothing TO scale! Too many people worry serving 10,000 users instead when they don't have any to begin with.

submitted 47 days ago • 0 comments

Tests or nah? I've worked with a few teams with no testing.... And every single time, development was fast at first. Then we hit a point when we'd have to spend so much time manually testing, shipping new features took forever. But so many teams don't want to write tests! Is this just me?

submitted 48 days ago • 0 comments

Last year I was an expert in NFTs and the year before that was crypto. So believe me when I say I know what I'm talking about when it comes to AI. /s

submitted 48 days ago • 0 comments

New business idea... "We fix all the security issues in your vibe code so you don't get sued into oblivion when software gets hacked."

submitted 49 days ago • 0 comments

Curious to know what your favorite part of being in tech is... Not the obvious stuff like remote work or compensation. Is it solving problems? Building stuff? Having to keep up with new tech? I personally like working on something, stepping back, and saying "I built that."

submitted 49 days ago • 0 comments

"I can get that done for a 1/10th of your price!" "Awesome. Send them my way and I might have some projects they can help me with." Or... "Then why are you talking to me?"

submitted 50 days ago • 0 comments

A large part of learning to code is having the courage and grit to try, fail, feel stupid, and then try again. Only following tutorials is like a safety blanket that slows your learning down and keeps you from feeling uncomfortable.

submitted 50 days ago • 0 comments

I admit it... I was wrong. I've been beating the drum that AI isn't really a big threat to software engineers (at least not yet). But I haven't been telling the whole story. What most people don't understand is that a software engineer's job isn't to write code. Sound odd? Stay with me here...

submitted 51 days ago • 1 comment

I see memes hating on how complicated front-end development has become. That you need a fancy Javascript framework, 20 node packages, etc... And "you can just write it in vanilla JS, HTML, and CSS..."

submitted 51 days ago • 1 comment

If I started learning to code from scratch again, I'd ___________. (thoughts?)

submitted 52 days ago • 0 comments

"Our new AI product is going to disrupt our industry." "It's a chatGPT wrapper isn't it?" "DISRUPT OUR INDUSTRY!" (In all fairness, the tech doesn't really matter as long as it solves a problem."

submitted 52 days ago • 0 comments

One of the biggest responsibilities of a tech leader or manager is to remove barriers for your team. Make sure the devs can focus on developing. Help the designers spend more time designing. I want my team to focus what they're good at - and what they like doing. Thoughts?

submitted 53 days ago • 0 comments

The first ever "software" I wrote was in Excel writing some functions and a macro. Was it ugly? Yes. What it low tech? Yes. Did it solve a business problem? Yes.

submitted 53 days ago • 0 comments

You don't need to know a a hundred different programming languages, frameworks, and tools to write and ship software that solves a problem.

submitted 54 days ago • 0 comments

Clear code is better than clever code.

submitted 54 days ago • 0 comments

New features or scalable code? How do you make the decision to spend more time writing better code vs pushing new features the marketing and sales team can sell? There's always a tradeoff. Curious to hear what you think - and what your title is.

submitted 55 days ago • 0 comments

"I'm not really sure what this code does." Then don't push it.

submitted 55 days ago • 0 comments

"You're an engineering manager? Must be nice to just sit in meetings all day, send slack messages, and get a paycheck." I do those things so my team doesn't have to. My goal as a manager to help my team get the resources they need so they can just code and not worry about anything else....

submitted 56 days ago • 1 comment

The cold hard truth about companies going all in with AI... Some tech leaders are really excited about replacing engineers with AI. Fewer engineers means lower development costs, right? For the sake of argument, let's say AI can replace engineers now (that's another discussion entirely...)

submitted 56 days ago • 1 comment