mrbonner

So I think the permissive license like Apache 2 without opening up for upstream contributions would be the way of future open source development. In a very similar way of SQLite. It definitely will help shield off a bunch of AI generated contributions first and foremost and then, aligning mental model with the human contributions with Mojo leader team.

Jeez, llama.c++ is becoming the ffmpeg cargo cult CLI now

I use Bonsai 27B ternary on my 24GB MBP. But, I believe you definitely can run it with 18 or even 8GB.

I have had 2 encounters with Dr Norvig, all online. One, I randomly reach out for him to get a referral to a hiring manager at Google. He gave me a couple of email leads. Two, I emailed ask him about a writing piece regarding fuzzy logic and why it was not popular in the US academic landscape. I got a very long, specific and detail response back. I need to print that email out and frame it lol.

I use coding agent but still don’t see the need to manage a swarm of them like this. Is there an actual use case for this or this is just for tokenmaxxing?

Hah tales as old as time. what’s next? Distillation of Disney theme park?

I agree with Odin creator here. I am not sure when or why we have come to a conclusion in our professional industry that "real software engineer uses vim/emacs/insert your CLI editor here" type of mental model. Is it just the testosterone level speaking here?

I read “how to win friends and influence others” many times but didn’t realize the main lesson is not to argue. That is until I reached 40. So, some lessons will take certain age to understand. I bet the OP is not in their 20s or 30s even.

Fun fact: it is spam filtering application that makes Paul Graham famous (and rich)

Then how do they get promoted?

This has been the the case all the time: most of our work are connecting pipes using glue code/duct tape so to speak. But now, we are delegating this to AI. Nothing new for me.

Maybe it is just me: I feel Anthropic most recent product announcements resemble more and more like what IBM tactic was at its high. For instance, the Watson AI hype after it defeated Kasparov. The difference is IBM actually wanted and let businesses buy and use Watson as opposed to time released like what Anthropic does to even boost the hype higher.

I really appreciate the documentation. But, it appears to me that this is how I also use Claude daily and I thought I am just using it as a coding agent. The intro however sounds like a recipe to use Claude for everything else beyond a coding agent.

Also, this stuff feels like alchemy to me . I bet some of you have the same feeling.

I’m doing it this year. Does the pill work as effectively as the drink?

I bought 3 programming books last month including this one. I really enjoy “Learning Go” over the official go book.

That just means there will be legitimate ways to short them.

So I have seen people using Claude to change variables’ names. Is it tokenmaxxing?

We have had a service to add two numbers. What make you think this is not realistic? :-)

good luck and take care of yourself!

Your method of combining models to strengthen the implementation reminds me of how we form stronger alloys by combining metals!

Don’t listen to anyone saying it is fine for reading or writing extensively with the xReal. I have one and it is PITA to do that over a long period. You better just stick with watching videos or playing games with it.

It’s all for show I guess. But at this point, why would anyone be surprised about it?

So this is the norm: quantized version of the SOTA model is previous model. Full model becomes latest model. Rinse and repeat.

Cool! I checked the source and noticed that even LLM prefers simplified, high level Rust coding styles: use value types such as String, use smart pointers such as reference counting, clone liberally, etc… instead of fighting the borrow checker gatekeepers.

It is the style I prefer to use Rust for. Coming from Python, Typescript and even Java, even with this high level Rust, it yields incredible improvement already.

Unverifiable software stack now amplified with LLM undetermistic. This while thing starts to feel like we are building on top a giant house of card!

You're talking about the aladeen or that aladeen? I don't understand which aladeen you are talking about.

This is great. I think Apple bought Kuzu, a in memory graph database in late 2025 to support RAG in combine with their FM like this. Even with such as small model, a comprehensive context of our personal data in graph RAG would be sufficient for a PA system. Do we know if we can have access to this RAG data?

Godspeed AI-I

Can you at least read the article before criticizing them? They explicitly call out that they use Bayesian Optimization (Gaussian process) thing for this. It is "AI" but not "LLM" like you think it is.

I am not sure if this is an April fool joke anymore in the age of AI.