HN user

robertritz

190 karma
Posts5
Comments35
View on HN

United States v. Heppner mentioned a public chatbot service. If a law firm (or specialized provider) offered a chatbot using their own servers and hosted the traces and other data on the law firms own servers it would almost certainly be protected. But another case would need to happen to determine that.

But that only applies for clients using the chatbot. If a lawyer is using the LLM it is definitely protected. No different if a lawyer searches something on Google or Lexis Nexis. The search itself is protected. I guess you could debate metadata but the content surely is protected.

MOOCs have always given value to a certain group (about 3% of people who are motivated to self learn) and are basically useless for everyone else. Which is why they are primarily virtue signaling devices to fluff your LinkedIn.

When I started learning to code for data science I appreciated the lectures and examples. Once you reach a certain technical level these MOOCs aren't really helpful though.

This I think will be the future. The ERP I made for my company is hopelessly tied to our process, and it saves time and reduces mental overhead for everyone.

It evolves as we do with me making feature updates and bug fixes every few weeks. Of course this is unusual in that I’m a very technical owner, but I feel this is the right way.

As software becomes easier to make custom software will reduce in price as well. Software gives my company an edge and I’ll take every advantage I can get.

I’m the owner of a smallish furniture manufacturer. About 15 employees. I built out the order management system myself because nothing really fit our process.

After looking at the site I can’t really say I know how this software could help us. I’ll look at it later on my desktop but first I think some better demo videos or gifs on the landing page would be nice.

I like the idea, but it's really just three summaries each day, which I can get from Google News. Would be nice to have links to the tweets at the very least.

Also the top three stories are most likely what your audience already knows. What people are craving, at least what I'm craving, is something that isn't on my feed everywhere else.

As someone who isn't particularly religious, but grew up in a religious household, and as someone married to a very religious person (different religions), I believe it's all about outlook.

Religion tends to give you several quite positive beliefs about the world that aren't entirely logical. Things like karma, the golden rule, belief in a plan, etc.

Generally speaking I also believe that religious people are more willing to trust and forgive. These are all pretty positive things.

And finally I believe religious people have a higher sense of duty to others, but the better term is probably responsibilism.

Meh I took a serious left turn after college and my first few jobs. Much happier now.

Work on Capitol Hill for less than a year then tech outsourcing then consulting. Realized it was boring, useless, and mind numbing and moved across the world. Now have multiple businesses, more than 30 employees across those business, and I get to have fun. It's stressful sometimes but I think we've kicked the stress finally (at 37). Now it's just fun and we get to see what we can pull off when we want to.

Most people simply quit or aren't willing to do the uncomfortable things. It's uncomfortable to be unbothered. But I certainly didn't follow my interests. I used my interests to get better at what was in front of us. Gotta pay the bills and give people what they want, I just put my own spin on it.

This recent article[1] from the economist shows how much of the aid spending around the world just replaces government spending and never creates a system where those countries can stand on their own.

Not always the case but definitely for a lot of aid spending. Where I'm located (Mongolia) aid spending just gives the government an excuse to free up the budget for handouts to get them elected. It's pretty useless.

[1] https://www.economist.com/leaders/2025/03/06/the-demise-of-f...

This exactly. Losing a few things pales in comparison to less mental overhead in every single task I do.

All I do now is keep a notebook with a rolling to do list where I make a new page every day and write what I have to do. Meeting notes also go there if they don't belong in some project files.

The only thing I wish I could do is keep notes from books somehow in an easily referenceable format. Kindle sucks, Obsidian is 180 from TikTok but still brain rot.

I gave it a softball question and it got it horribly wrong.

"Cidr /17 is what subnet mask?"

Here is it's answer:

"A simple subnetting question!

Given: CIDR /17

To determine the corresponding subnet mask, we need to understand how CIDR notation works. CIDR (Classless Inter-Domain Routing) notation is a compact way to express the relationship between an IP address and its associated routing prefix. The number after the slash (/) represents the number of leading 1-bits in the subnet mask when it's converted to binary.

A subnet mask for IPv4 is 32 bits long. Here's how to convert /17 to a subnet mask:

Determine the number of 1-bits: /17 means the first 17 bits are 1. Determine the number of 0-bits: Since the total length is 32 bits, and 17 are 1-bits, the remaining bits are 0. So, 32 - 17 = 15 bits are 0. Convert to binary and then to dotted decimal: Binary: The first 17 bits are 1, and the last 15 bits are 0. 11111111111111111 000000000000000 Dotted Decimal: Divide the binary into four parts (each 8 bits long, known as octets) and convert each to decimal. 11111111 = 255 11111111 = 255 00000000 = 0 000000000000000 (last 8 bits of the 15 zeros) = 0 So, the subnet mask for CIDR /17 in dotted decimal notation is: 255.255.0.0."

Yes it's a simple subnetting question and it got it wrong. Right answer is 255.255.128.0.

MLX only works for fp16 right now. If it ever works quantized I will almost certainly move my app over to MLX instead of llama.cpp.

My app also uses a very small (30MB) PyTorch model and shipping it requires an extra 100MB for PyTorch in the app. Very very stupid.

I think its important to remember that last mile inference is still pretty bespoke for most things. If we want to see gen AI stuff take off and now have the big cloud providers in charge this needs to be fixed.

Apple is in a good place to solve at least part of the equation.

Yes! 1 consistent global universe would get me to jack into whatever matrix they were selling.

It could even be multiple games connected with sufficient break in context. Eve for space and another game for ground assault. But they actually interact live.

Then throw in next gen VR and you basically have ready player one.