highmastdon

The things mentioned in the post aren’t code. That’s why code isn’t the hard part. Writing the algorithm to x y or z isn’t the hard part. Knowing that the algorithm is needed, it’s beneficial, there’s no alternative and the optimization isn’t in vein, that’s the hard part.

Structure, design patterns, software architecture, systems engineering, optimization (not of the LoC but the system), data accessibility (in structure as in implementation like indexes etc)... those are the hard ones that require engineers to design and implement properly.

And that’s only going to be more as it’s not the 20 medior SWEs that throw together a full stack application in two days but the untangling of AI slop as the architecture wasn’t built for hundreds of concurrent users

This list is so absurd.

We’ll face a bigger problem sooner rather than later, which is population collapse. Youth don’t procreate amy more. Birth numbers are at an all time low. We see the issue arise in rats and the experiment is all too relevant for the current age of social media and fearmongering (John B. Calhoun’s rodent “utopia” experiment). All these “Grand Challenge” problems seem trivial to that.

Furthermore: Why is 1 here when 2 is present. Again, solar is usually not relevant when we want power when it’s dark… even theoretical it wouldn’t work. We’d need a high capacity dirt cheap storage, and even then we can’t keep it till winter when there’s no sun to go around and effectively supply 3. Irrelevant when there’s population collapse. And even then, why would we want this instead of reforestation and low depth water protection from fishing and environmental issues.

5. How is this even a problem. Unless we got corrupt(ed/able) governments (read: lobbies) that allow exemptions in every law designed to protect the environment (also, settlements are a twisted way to fill governments pockets instead of rooting out evil) 7/8/9 the inverse effects are even worse health, as everything is fixable. 9 would incur even more social isolation, more so than the internet did 10 seems to be the first that’s actually reasonable Same for 11 For 12 see 9 13 yes but that’s something that a self learner would already be able to do. AI is at a level that we can manage

/rant

Speak facts. Don't hide behind morality this, care that, environment and virtue signalling... Tell me, what is the factual things Grok model does wrong and is factual incorrect?

1. intelligence to token-use ratio is much better than the frontier models 2. hallucination rate is MUCH lower than OpenAI, Google and Deepseek models 3. the cost for output tokens is much lower than Anthropic models while near same intelligence and task completion 4. task completion time and success is on par with Opus 4.8 Max, but 3.5x less time used

Especially 2. you're probably more willing to allow OpenAI to bankrupt your morality, than face the facts of the model itself

Overall, Grok models are more factual and less politically influenced than OpenAI ones.

I’m using MacParakeet these days. If your language is supported, definitely give it a try. It’s much faster and lower footprint

Fyi, your bottom bar with links isn't responsive causing annoying behaviour with horizontal scroll on mobile phone

Still to this day I don't understand how people manage to do software development on a Windows PC. I started with my first secondhand MacBook which was one before the unibody, and then upgraded to unibody and never have left ever. The only reason I was forced to use Windows was for clients that didn't have support for MacOS or Linux. The only thing I need Windows for is rootkit enabled games... which pains me, otherwise I'd have Linux on my desktop

I had the feeling that Cerebras only supports smaller modes. Maybe something to do with their hardware arch? I never dove into it. I wanted to use Kimi K2 fast for coding and Groq was the only fast provider at the time

The nice thing is that adding this to the basic prompt that cursor uses will advance all those users and directly do away with this problem only to discover the next one. However, all these little things add up to a very powerful prompt where the LLM will make it only easier for anyone to build real stuff that on the surface looks very good

Don't underestimate the amount of knowledge you don't have to perfect things. I remember building scheduling software in PHP in high school, because I just fixed problem after problem, and I was not limited by any form of knowledge. If I'd have to do it again, I'd be perfecting the architecture, refactoring everything every other week...

Why wouldn't they redistribute to other maintainers of the plugins/clients. They have much more knowlegde of where money is needed or where resources are lacking.

Or pay devs, not in their circle to build the features the community wants but they can't (lack of expertise, or otherwise) implement?

Would this be something to use over Docker containers when running a single host homelab server? What would be the cons/benefits?

Didn’t Google’s closure compiler so something similar way before we had imports?