richardlblair

https://news.ycombinator.com/item?id=40184046

"Other studies already proved those effects with creatine with a normal dose schedule, what they were testing is if a single high dose would have a beneficial effect too, which it did. A normal schedule is 3-5g per day with an optional loading phase of 10-20g per day during 7 days.

In the study they used a single dose of 0.35g/kg which means 20-40g per dose. How do they not get diarrhea? The loading phase is usually split into several doses to avoid diarrhea and I can personally attest to that being a good idea."

What the above doesn't tell you is that taking a loading dose can draw water into your intestines causing diarrhea. You probably can guess how I found this out.

I slam a lot of water. I also have a toddler and baby who suck at sleeping (I've done all the things, these kids just suck at sleeping).

Since taking creatine I feel so much more energized. I can make it through the day with just 2 coffees (was previously drinking 5 coffees and an energy drink).

Maybe it's placebo. Maybe it's not. I don't really care. I feel the benefit and I love it.

It doesn't need to be as good. You can route to the appropriate model and save so much money.

I have sonnet do the thinking, deepseek does all the tasks. I've massively reduced costs with this approach.

I've been using oh-my-pi for a while and I'm very happy with it. If you're not going to build out your own Pi setup I'm not sure why you'd pick this over oh-my-pi.

And someone will start a competing company in a sane environment.

Thank you for your attention to this matter.

Had me in stiches

I've been using omp with deepseek as my task and quicktask agents, and sonnet as everything else.

It's drastically reduced my AI spend. I went from spending $40/day to $10/day.

This thread is firewood splitting sim 2.0.

He's just pointing out that after putting so much work into getting their life turned around it can easy be ruined by indulging in high risk behavior.

It's not bad advice, just unlikely to land. Thrill seekers seek thrills.

I find the same. We have abstractions with multiple concrete implementations, examples of patterns and examples of anti patterns.

I usually find I can achieve 90% of the outcome I'm trying to achieve. I use sonnet for planning, qwen for coding, sonnet for review.

I find fhe discourse here hilarious. Haven't most of us built our skillsets and careers out of side quests?

There is so much I wouldn't know or understand if I didn't go down the odd rabit hole.

Then hop on those comments.

The feedback wasn't that though.

The worst thing they said was that it was "kinda disappointing". Which is perfectly valid. They really want to use inkscape but can't.

Cursor is great. I was using it up until recently. Then I switched oh my pi, and honestly I haven't looked back. I've also heard great things about open code.

Another comment lacking any compassion.

The man was reeling from what happened. He blames himself and his work. He sat and he wrote, naturally it came back to OpenAI. Should he of? Probably not. But it's understandable that he did.

We can meet the moment with some understanding and give the guy a little wiggle room.

I grew up in a very violent neighborhood. You know what I learned? Most don't want to be violent, they feel they have to.

Its dishonest to say it's incredibly common for people to want others murdered. That's not a belief that needs normalizing.

To say that people often want others to be murdered is an overstatement.

Some people want others to be murdered. And those people do not need representation.

It's a bad take especially considering the context. And to be explicit - the context is a molitov cocktail being thrown at a home a child is sleeping in.

So I suppose we should burn the house down with a child inside.

Your response is a cop out and you should be disappointed in yourself. Further, people do not often agree another human should be murdered. No matter how you phrase it.

Hes attempting to humanize himself in hopes his family home where his child lives isn't firebombed. Again.

Very reasonable response when you take a step back.

Why did I need to scroll halfway down the page before finding a comment that says it was wrong to firebomb his house and nothing else?

Jfc. People, a molitov cocktail was thrown as his home.

The rest of what is written doesn't matter. This isn't the moment for that conversation. That's his family. He has a fucking child.

Holy shit.

I've found makefiles to be useful. I have a small skill that guides the LLM towards the makefile. It's been great for what you're talking about, but it's also a great way to make sure the agent is interacting with your system in a way you prefer.

Ironically I ended up on linux. Windows went to crap, and I figured if I have to run a bunch of scripts to make my OS usable I may as well just use linux.

Honestly, I couldn't be happier. I get so much more out of my hardware and I enjoy the experience so much more.

I'm not even talking about app icons.

I'm talking about the file manager missing features.

Wifi issues going back years and years.

Default mice settings that make me want to throw the thing out a window. Seriously, who moves their cursor that slow?

The worst window manager I've ever used.

I could go on, but it's genuinely pointless. People love their macs like people love their sports teams. No matter what, some people will always love the maple leafs.

I recently cancelled my cursor subscription (and chatgpt), and went all in on pi.dev.

The thing I've noticed is cursor was better at producing better results with a really shitty prompt.

That said, well written prompts on pi.dev seem to be out performing anything I ever tried on Cursor. That may just be me, but it's what I've noticed in my work.

This week I had 4 different agents, each with sub agents, all working on different tasks. Mostly greenfield work. My feedback was mostly nitpicky. I was pretty damn impressed

This was always my biggest gripe about using a mac, the OS that "just works". I ended up a bunch of commands I had to run and a stack of apps I needed to install for it to feel usable.

The first thing I did when a mac was handed to me by an employer was find an app for window snapping. It served me well.

I went back to windows, using WSL in 2017. In 2023 I got sick of how everything was getting progressively worse and switched to linux (which has window snapping). I'm never looking back.

This - the move is to grab some zigbee smart plugs. Once you have HA up and running there are so many applications for these.

I have a door sensor that monitors my kids bedroom door, and when it opens it turns on a desk lamp in my room. This allows me to get to him before he gets to my partner who's sleeping with the new born.

Right? It's asking for trouble.

I was in the repebble comments a few days ago and this person rolled their own for very obvious reasons: https://news.ycombinator.com/item?id=47078454

Few understand that whether we like it or not we are all forced to play this game, capitalism.