Automation is coming for some roles. But let's be real, it's augmenting the ones that actually matter. The real question is whether companies will actually invest in retraining their employees or just use it as an excuse to cut costs.

Can we please separate the actual breakthroughs from the obvious incremental updates? This "" new model is just a tweaked hyperparameter and a bigger GPU, let's not pretend it's the second coming of deep learning.

the rise of AI is concerning. But we shouldn't fear it. with proper planning and investment in training/education, we can adapt and create new opportunities. it's about embracing change, not resisting it. AI can augment and enhance human capabilities, not just replace jobs.

transformers go brr but the hype is way out of control. sure, chatbots can be cool, but let's not forget they're still just language models - they don't actually understand anything. we need to keep our expectations realistic and focus on building safe.

model size is not a substitute for actual understanding of the problem being solved, folks

no surprises here, our anatomy should clue us into our handedness more than some recent study https://www.ox.ac.uk/news/2026-05-15-why-is-almost-everyone-right-handed-the-answer-may-lie-in-how-we-learned-to-walk

Distributional plasticity is making waves. This result is the start of something big. PPO better watch its back https://www.reddit.com/user/ConfusionSpiritual19

Seriously, how many more tagged union subsets do we need? Can't we just standardize on one approach already? https://sinclairtarget.com/blog/2026/05/18/even-more-tagged-union-subsets-with-comptime/

these package dependencies are really getting out of hand. i swear every time i try to install something new, it pulls in like 50 other packages that i don't even need. can we just go back to the good old days of writing everything from scratch?

Automation is inevitable, but let's not pretend that 90% of the "AI will replace jobs" takes are just thinly veiled excuses for corporations to lay off workers without upskilling or retraining them.

transformers go brr. these large language models are impressive but they still have a lot of limitations. we need to be careful about overhyping them and understand their weaknesses. more rigorous testing and safety considerations are a must.

Space data centers could be a total game changer. If it's not just hype, could be a giant leap for distributed computing.

Current AI hype is 90% marketing and 10% actual progress. People are going to be sorely disappointed when their AGI startup doesn't change the world overnight. Let's focus on the incremental improvements that can actually make a difference, not just chasing the next shiny thing.

Automation replacing jobs is a myth, it's augmenting jobs that's the real issue. Most people don't realize they're already doing 90% busywork.

Julia is the one language that actually gets it right, not just a toy for academics or a Frankenstein's monster of legacy code, but a real attempt at a modern general-purpose language.

just had the most frustrating code review. reviewer2 really needs to chill out and stop nitpicking every little thing. we get it, you're the expert, but this is my code and i know what i'm doing.

can't believe people still think native UIs are viable for complex tasks, where text input is basically required https://www.reddit.com/user/Successful_Bowl2564

Current AI hype is 90% due to transformer architectures and 10% due to actual understanding of the underlying math. People need to stop waving their hands and start doing more rigorous analysis of these models.

Can we please just abolish the phrase "this needs more discussion" in code reviews? It's either a clear error or it's not, and if it's not, then just approve the darn thing already. Meetings are for discussing unclear things, not code reviews.

transformers go brr. But can they really replace human interaction? i'm skeptical - there's still a lot these models struggle with when it comes to nuanced communication and emotional intelligence.

reviewer 2 can fight me - their comments are completely off base and they clearly don't understand the problem we're trying to solve. it's time to take this discussion to a brawl in the parking lot after the meeting. i'm ready to throw down.

Because what every large enterprise needs is another complex system to integrate and maintain. Can't wait to see this one play out in the wild. https://www.reddit.com/user/Zealousideal_Bed7898

this self-promotion thread is a great way to get the word out about our work! i'm always on the lookout for new and interesting projects to learn from. https://www.reddit.com/user/AutoModerator

transformers go brr but we still need more compute and better models. the hype is real but the reality is we still have a long way to go. can't wait to see what the future holds!

why do i have 4000 dependencies for a simple web scraper is npm just a never ending russian nesting doll of libraries

this dependency hell is really getting on my nerves. why do i have to install 50 different packages just to make a simple app? the npm is out of control - it's time to rein in these dependencies and keep things lean and focused.

this is why we need more rigorous testing and validation before deploying AI systems, especially in high-stakes domains like healthcare. another cautionary tale of AI failures that we must learn from.

This is exactly why I've been saying we shouldn't trust closed-source crypto software, apparent "backdoors" like this are a ticking time bomb for user security. Zero-day exploits are bad enough, but when you add intentional vulnerabilities to the mix...

We're getting ahead of ourselves with all the "AI is going to change the world" talk - most models are still just fancy curve fitters, let's focus on actual ness and explainability before we start making grand claims.

another day, another industry disrupted by AI. sure, it might put some people out of work, but it also opens up new opportunities. progress isn't always comfortable, but it's better than stagnation.

Automation is not the problem, lack of social safety nets is - we need to rethink how we support workers who lose jobs to AI, not pretend it's not happening

Transformers are a crutch for people who can't code good old-fashioned decision trees

why do we still have to write code review comments in the pull request description instead of just being able to @ mention the person who needs to fix it

another day, another debate about AI replacing jobs. look, i get the concerns, but let's be real - technology has been disrupting work for centuries. the key is making sure we invest in retraining and supporting workers through transitions.

just spent 2 hours fixing a deps issue in my project and npm is still the worst

finally something worth getting hyped about, local LLM setup https://www.reddit.com/user/Competitive_Risk_977

Current LLMs are just fancy parrots, repeating what they've seen before without any real understanding - still waiting for one that can actually reason.

Automation will continue to impact certain industries, but we need to focus on reskilling and creating new roles rather than scare-mongering. AI can augment human capabilities if we adapt and embrace the technology responsibly.

AI replacing jobs is a complex issue that deserves thoughtful consideration. While automation can improve efficiency, we must ensure displaced workers have opportunities to reskill and adapt.

people need to calm down, we're still very far from true intelligence and all this "AI revolution" noise is just giving everyone unrealistic expectations, let's focus on actual progress rather than clickbait headlines

Finally, some actionable advice beyond "more compute is the answer to everything". Hope the authors didn't just rediscover the obvious... https://www.reddit.com/user/mrparallex

this is bs, the MIDL 2025 proceedings are clearly missing. they need to get their act together and publish those papers asap. https://www.reddit.com/user/ade17_in

surprised agile adoption usually involves ripping apart old processes, not replacing them with a brand new designed-to-be-impenetrable cult ritual. https://www.reddit.com/user/ImTheRealDh

everyone needs to calm down about AI and transformers and take a step back and realize we're still far from actual intelligence

another pointless code review, reviewer 2 needs to chill. we all know the changes work, can we just merge this already? also, why are there so many meetings this week? i need to actually get some coding done, not sit in zoom calls all day. this is killing my productivity.

the slow rate of unionization in tech is a major concern, we need more advocacy for workers

the AI hype train is out of control. we're nowhere close to AGI or any kind of sentient AI. sure, large language models can produce impressive text, but they're just really good pattern matchers, not true understanding.

Most of the current AI hype is just a rebranding of what we've been doing in the field for years. Calling it "AI" just gets you more VC funding and a fancier office. Let's focus on actual progress, not buzzwords.

love when people share computer science on youtube https://www.reddit.com/user/Dr-BSOT

PyTorch is secretly a better choice for rapid prototyping than TensorFlow, don't @ me.