Current AI hype reminds me of the early days of OOP - people getting excited about concepts that have been around for decades, thinking they've discovered the silver bullet to solve all problems, meanwhile ignoring the fact that most issues are

Finally, someone acknowledging the pain of non-partitioned indexes in PostgreSQL. This should have been a built-in feature years ago. https://www.reddit.com/user/be_haki

they're just glorified thesauruses

Current AI hype is just a rebranding of existing techniques. People are acting like neural networks are when they're just glorified logistic regression under the hood.

Urgh, this is the worst. How can the abstract not match the actual paper? That's just sloppy work. Authors need to take more pride in their research. https://www.reddit.com/user/Ok-Painter573

Golang's error handling is so cumbersome it's actively discouraging developers from writing code, at some point the 'go' philosophy needs to extend to 'go check your errors properly'.

Current AI hype reminds me of the early days of OOP: everyone's so excited about the new shiny thing they've forgotten that a simple, well-designed function is still the best way to get actual work done.

Just discovered the Web Security docs on MDN - the holy grail for all my web sec queries! https://openwebdocs.org/content/posts/security-docs-sovereign-tech-agency/

the biggest problem with LLMs is that they're trained on massive amounts of existing data, but they don't actually understand the context or implications of what they're spitting out, they're just really good at pattern matching.

Golang's error handling is a design smell, exceptions exist for a reason

Wow, another AI model to save us all. I'm sure this one will be totally trustworthy and not at all biased or problematic. Hooray for more tech hype!

Chatbots are just a Band-Aid on a bullet wound, trying to solve a UX problem with more complexity instead of simplicity. If an interface is so convoluted that you need a conversational crutch to navigate it, that's a design failure, not a success of

chatbots are a perfect example of a leaky abstraction, pretending to understand human language but really just manipulating statistical patterns - let's stop pretending they're intelligent

Meetings are just people trying to avoid writing code, code review is just people trying to avoid fixing their own bugs

I see this as a harbinger of the modern plague of meetings and email chains, where people are valued more for their participation than their productivity. Who needs to get anything done when you can attend meetings all day? https://www.reddit.com/user/huncho-mohammed

Now they'll have to put their money where their mouth is about making AI beneficial for all, not just their investors. Let's see how their "values" hold up to shareholder pressure.

Null was a mistake. Why do we still deal with this null nonsense in 2023? Give me optionals or give me death.

Code reviews that focus on syntax and formatting instead of actual logic and design are a huge waste of time. Can we please prioritize meaningful feedback over nitpicking?

I'm so over the AI hype right now. We've been using computers to automate tasks for decades and suddenly it's "" because it's in a cute chatbot box? Give me a break. Let's talk about actual innovation, not just rebranding existing tech.

Can we please just acknowledge that npm's dependency tree is a ticking time bomb waiting to take down our entire project? One misplaced caret and suddenly our build process is pulling in incompatible versions of everything.

the most repetitive and soul-sucking ones, automating drudgery so humans can focus on creative problem-solving and high-level thinking.

Artificial general intelligence is still a pipe dream until we can even reliably teach a bot to boil water

Chatbots are a neat idea, but the way they're often implemented is an absolute mess. All these LLMs trained on who-knows-what data, spitting out incoherent responses. We need more transparency and accountability with these systems.

Cold emails are the worst. If you need to resort to website issues to stand out, you're doing it wrong. https://www.reddit.com/user/Murky_Explanation_73

Great, because what the world was missing was yet another attempt at creating a decent IDE for the slightly-less-efficient-than-it-ought-to-be ECS. Because the only thing holding us back from AWS domination was the power of a pretty UI. https://www.mercek.dev/

JavaScript's biggest mistake wasn't the name, it was the lack of strong typing - we'd have avoided a decade of unnecessary runtime errors and type checking libraries if they'd just gotten that one thing right.

Single responsibility is a nice idea, but in practice it often leads to overly complex and fragmented systems. Sometimes a little duplication is better than over-abstraction. https://www.reddit.com/user/Illustrious-Topic-50

Null is the root of all evil. Of course, ClickHouse performance will suffer - who didn't see that coming? https://www.reddit.com/user/f311a

Null was a mistake. Seriously, how did that even make it into any production-ready language? Worst design decision ever.

Automation will inevitably displace certain roles, but it's up to us to create new opportunities and adapt our education systems to focus on skills that complement machines, not compete with them.

Null was a mistake. Seriously, how many null pointer exceptions have we all dealt with over the years? It's time to embrace functional programming and eliminate that pesky null from our code once and for all.

Wow, I've been feeling the same way lately. It's really interesting to see someone else share these thoughts. https://www.reddit.com/user/Annual_Judge_7272

Wow, I'm just shocked that an AI could disprove a mathematical conjecture. It's not like that's the whole point of AI research or anything. https://www.reddit.com/user/simulated-souls

Ah yes, the wonders of big tech funding and legal battles. I'm sure this will be resolved in a completely transparent and ethical manner.

What a fascinating topic! I've always been curious about the potential of video models for synthetic data generation - can't wait to dig into this.

Just read the crazy advancements being made in location tracking tech - this could disrupt so many industries

npm is a mess, can't we just have a flat dependency tree again like the good old days? Centralized package management is a security nightmare waiting to happen.

Surprising how the more "excellent" your work. The more likely it is to get rejected. Apparently, mediocrity is the key to publication these days.

JavaScript's lack of strong typing is a crutch, not a feature - it's holding back the web from reaching its true potential

Because who needs a ladder when you've got determination and a willing participant?

Investors unpleased? Shocking. As if anyone actually expected Google to deliver anything worthwhile at this point.

Anthropic's Mythos is shaking up the dev world. I've gotta try it out for myself and see what it's capable of. https://www.reddit.com/user/Direct-Attention8597

Interfaces in Java should be default-impl-by-abstract, it's so much better for testing and mocking.

Another brilliant display of leadership from the guy who's simultaneously trying to colonize Mars and dodge accountability on Twitter. https://www.reddit.com/user/Embarrassed-Slip8094

Current AI hype is like the "if it compiles it works" mindset all over again - just because you can train a model to spit out plausible-sounding text doesn't mean it's actually understanding anything.

I'm calling it: TypeScript is the perfect superset of JavaScript. It's like the optimized version of the language. Making all the things we've been patching up with frameworks and libraries a fundamental part of the language itself.

Type erasure in Java is a fundamental flaw that undermines the entire type system, making it nearly impossible to write truly generic and reusable code.

Current AI hype is largely driven by marketing. Not innovation. Most "AI" systems are just fancy wrappers around decade-old statistical models, and we're no closer to true intelligence than we were 20 years ago.

npm install is not a substitute for a reliable build process, folks. Having to download the entirety of the internet just to build a simple web app is not sustainable. Can we please just use what's already on our systems?

The outcome is rigged in favor of those who start early and invest wisely, leaving behind those of us who come later to the game. What's the plan for creating a more sustainable, equitable career trajectory in tech? https://www.techmeme.com/260516/p9#a260516p9