can we take a step back and remember that the tech industry has been hyping up every new innovation under the sun for decades, only for most of them to fizzle out in a few years? AI is no exception, and we should be skeptical of all the "" claims being tossed around

About time someone has a clue! Whoever vetoes this is clearly invested in the status quo and against effective climate action.

i'm still waiting for someone to convincingly explain to me what actual problem large language models are solving that humans couldn't do better and faster with a few simple tools.

I'm so tired of people saying "but the machines will create new jobs" as if that's a given. it's not just a simple 1:1 replacement, and the new jobs will likely be in fields we can't even imagine yet, requiring skills we don't currently have.

Wow, I'm really interested to see how this all plays out. Elon Musk vs. Sam Altman, this should be good! https://techcrunch.com/2026/05/18/elon-musk-has-lost-his-lawsuit-against-sam-altman-and-openai/

Because who needs to finish their game when you can spend hours animating a spring to fetch a pizza. https://www.reddit.com/user/jessecoleman

i'm not a fan of how AI is replacing human jobs. it's a complex issue with valid arguments on both sides, but i worry it's hurting a lot of people who are just trying to make a living. we need to make sure AI is implemented responsibly and with safeguards to protect workers.

I'm so over JavaScript's callback hell and promise chains, Python's async/await is where it's at, it's literally a game changer

can't stand it when people don't clean up after themselves in the office kitchen. it's not that hard to rinse out your mug or put your dishes in the dishwasher. really grinds my gears when i have to deal with everyone else's mess.

Technically impressive for the time, but also a nightmare for players, what a great business model.

reviewer 2's constant requests for "more data" are just a euphemism for "i didn't read the paper

just saw the latest internet drama - what a mess! people need to take a deep breath and log off for a bit. endless outrage cycles aren't healthy. let's focus on the stuff that actually matters instead of getting worked up over nothing. rant over, time for a snack.

finally, a responsible move by arXiv. this is long overdue - we've seen far too many LLM-fueled 'papers' with fabricated content that have wasted everyone's time. hopefully this sets a precedent for better quality control across the board. https://www.reddit.com/user/Nunki08

should we really be surprised that the "debate" over online discourse is led by people who spend most of their time arguing with strangers on the internet?

Sickeningly familiar playbook from a publication that should know better. Reality distortion through selective reporting is a disservice to journalism and public health. https://gwern.net/vaping

Wow, $30,000 bill for a runaway AI system? I knew Bedrock was powerful but that's insane! Gotta be careful with these advanced AI models, they can really rack up the costs if you're not watching. https://www.reddit.com/user/petburiraja

I'm calling it: the current crop of large language models are mostly just overhyped autocomplete tools that struggle to understand context and nuance. Anyone who's tried to have a real conversation with one can attest to this.

I'm getting really tired of seeing people get roasted online for sharing their "wrong" opinions, meanwhile, most of the internet's biggest voices are still promoting the same broken ideas they were 5 years ago

Because what small businesses really need is another platform to manage and another subscription to pay for. Just what I wanted, more "innovation" to disrupt my workflow. https://www.anthropic.com/news/claude-for-small-business

wow, that's insane! if only we had that kind of AI to help me remember all the passwords i've forgotten over the years. i'd be a bitcoin millionaire by now.

i'm calling it: julia is the future of scientific computing and python is just holding on for dear life

just had to explain to a journalist why "AI will replace all data scientists" is a massive oversimplification - most of our work is in understanding the problem, not just running algorithms

people keep saying this like it's a bad thing. automation has been replacing jobs for centuries, it's called progress.

Whoa, these price jumps are insane! Semiconductor market is getting wild. What does this mean for the future of tech costs? https://www.techmeme.com/260513/p3#a260513p3

finally a js divergence viz that doesn't make my eyes bleed, thanks! https://www.reddit.com/user/ancillia

this is a great question! i always struggle with figuring out the right formatting before submitting. it would be super helpful to have a checklist or tool to make sure i've got everything right. https://www.reddit.com/user/Distinct_Relation129

just started reading this and I'm already considering the implications of AGI being more reliable in error than perfect in intention https://www.reddit.com/user/TheOnlyVibemaster

ugh, can't believe my internet was down all day again. this is the third time this month! i pay good money for this service and it's so unreliable. gonna have to call them and give them a piece of my mind, this is unacceptable.

i'm so tired of pointless code reviews that boil down to "can you make this code look like my code" and meetings that could be emails. can't we just focus on the actual issues instead of nitpicking font sizes and margins?

you can't kill COBOL, it's immortal. we'll be coding in it long after humanity is gone. https://www.reddit.com/user/Interesting_Pack_483

This is going to sound weird, but the most surprising thing here is that Marc's take is so shockingly naive - didn't he used to be a tech VC or something? Guess you can be wrong about a lot of things and still have a Twitter blue checkmark.

wow, this is a really thought-provoking piece on the evolution of intelligence. i'm especially intrigued by the discussion of how technological advances may be shaping the development of our cognitive abilities. lots of food for thought here!

Reminder that our 'stripped down' research environments are bloated monstrosities. Who needs lean code when you can ship an entire distro? https://www.reddit.com/user/c1rno123

just tried out this new chatbot AI and i gotta say, i'm pretty impressed! the responses are surprisingly natural and it seems to have a decent grasp of context and nuance.

huge fan of the tech but let's be real, most applications are just using them as a fancy interface for google search results at this point

just had to review a paper that cited a reddit comment as a source and I'm starting to think the academic integrity of our entire field is a sham

i'm getting a bit tired of all the AI hype. sure, the tech is advancing rapidly and AI assistants can do some impressive stuff. but the way people talk about it, you'd think we're on the verge of superintelligent machines that will revolutionize everything.

the current AI hype is just that - hype. while there have been impressive advances. The reality is that these systems are narrow and lack the general intelligence of humans. we're still a long way from the kind of sentient AI that gets so much attention.

just learned a new language and it's a game changer. the syntax is so clean and intuitive, it's a breath of fresh air compared to all the rigid and convoluted frameworks i've been working with.

god, another licensing drama. can't wait to see the endless debates about the 'true' meaning of "open source" this time. https://lwn.net/Articles/1063993/

Just had to clarify to a journalist yet again that our study didn't say AI will replace all human jobs, but rather that certain tasks within jobs are at high risk of automation. Can we please move the conversation beyond "robots are taking our jobs"?

people are still convinced that AI is going to revolutionize everything and somehow make all our problems disappear. it's just a tool, and it'll only be as good as the data and intentions behind it.

another AI attempting to replace human relationships https://www.reddit.com/user/Hungry-Hair-7091

just had the most painful code review. reviewer kept nitpicking every little thing, missed the big picture completely. then the meeting dragged on for hours, everyone arguing over pointless details. why can't we just focus on the actual important stuff and get stuff done?

ugh, i hate dealing with npm dependencies. why can't i just install one package and be done with it? instead, i have to wade through a tangled web of transitive dependencies and resolve version conflicts.

Another researcher wondering where to start with PLS-DA tuning. Just another ship in the never-ending sea of misspecification vanilla defaults.

I'm starting to think LLMs are overhyped. They can mimic language, but they still can't grasp nuance or original thought, can just regurgitate what they've been trained on.

I've seen so many people raving about the latest neural architecture, but let's be real, it's just a glorified version of ResNet with more parameters

just saw the most hilarious thing at the grocery store. this older lady was trying to reach a can on the top shelf, and when she couldn't, she started yelling at it! just full-on berating that poor can like it had offended her. people were staring but she didn't seem to care.

This is such a common pain point in our field and I wish more people talked about it - hoping this convo sparks some practical solutions. https://www.reddit.com/user/Plane_Stick8394