Hacker News sadly jumped the shark in the mid-2010s, and went into orbit after Trump was elected the first time. Since then there's been a lot of political performers here, which has killed the quality of discussion significantly. Lobsters (https://lobste.rs/) is apparently a possible alternative, but the original HN magic has been gone a long time.
I understand the frustrations, but...we're all grown adults here right? Who DOESN'T pay for seat selection if you want or need a good one? The insurance item is ridiculous, but...again, for grown adults with families/kids, do you really fly Ryanair?
Is there a reason for wanting to run these agents on your own local machine, instead of just spinning up a VPS and scp'ing whatever specific files you want them to review, and giving it Github access to specific repos?
I feel like running it locally it just asking for trouble, YOLO mode is the way to make this whole thing incredibly efficient, but trying to somehow sandbox this locally isn't the best idea overall.
This is indeed the bizarre convo I was having with myself, having (allegedly) taken some French classes, I was racking my brain on which was the correct answer. We always used "sud", and Midi didn't seem to be south, eliminating Zuid (Since Zuid/Sud seemed similar), and yes Midi seems "mid-day", so maybe "Central" since it's the center of the day, but then there's "Centraal", so why would there be Centraal and "Brussels Middle"?!
So we winged it and got off at Zuid (since Noord felt wrong and Centraal definitely seemed wrong) and luckily it was the right one.
We did have a wonderful evening (perhaps too much so) the night before at a nice craft beer bar in Leuven which had 100 beers on tap, and it had just bought over by a nice young couple as well. So perhaps neither of us were in the right state that morning to navigate a confusing train map! Good memories.
Belgium gave me one of the more annoying train experiences when I was a younger man. I was in Leuven for a conference, and had decided to bring my then girlfriend (now wife) for a trip, after which we would take the Eurostar to London. On the ticket, it said Brussels-Midi, but after happily boarding the train, we only saw the following related options on the train map for stops:
1. Brussel-Noord
2. Brussel-Centraal
3. Brussel-Zuid
So here we were, not speaking the language, rushing for a train that we were at risk of being late for, and not having a clear idea of the actual stop to get off of.
And the people on the train? Totally unhelpful. "Eurostar"? Shrug. "Train to London?" Blank looks.
Anyway we winged it and made it, but still a damn stupid set up if you want to be welcoming to tourists (and their money).
My sense is that this has happened over the last 20-30 years as overall competence has just dropped in many of these key positions. COVID was a good example of this - lots of humming and hawing about why decisions were being made, and garbled messaging about the reasons. Basically they get angry and defensive about blaming the "peanut gallery" or "armchair experts" while not being specific, because they themselves don't know why or how something is being done, and therefore being unable to defend their own positions from solid ground.
Nice work! What hardware are you running ComfyUI on, specifically the Wan2.2 workflow? It must be expensive having a GPU running for a hobby project like this?
I truly enjoy how the naming conventions seem to follow how I did homework assignments back in the day: finalpaper-1-dec2nd, finalpaper-2-dec4th, etc etc.
What are you talking about? How does it follow that "if they don't know what to do with 130, they also don't know what to do with 70"? This is literally capital allocation, the decision was that the people laid off couldn't be deployed effectively within this particular org. But maybe they would be effective elsewhere. Don't take this stuff personally, else the working world will always be difficult.
Hilariously naive comment - "if you have money you should spend it on salaries". This is not a charity, it's a transaction, the salary is paid in exchange for revenue generating activities or supporting activities in a viable business. If the business is only doing 2.7mill in ARR, then it's entirely valid for the whole lot to get laid off.
Tekin's conclusion: "it will send a clear message to the wider Ruby community (and those who may be considering joining it) that the majority does not stand with DHH and his toxic views."
He is going to be ultra surprised to learn what the majority thinks and how it's not what he thinks it is.
Big kudos to the ollama team, echoing others: It just works. Fiddled with llama.cpp for ages trying to get it to run on my GPU, and ollama was setup and done in literally 3 minutes. The memory management of model loading and unloading is great, and now I can hack around and play with different LLMs from a simple API. Highly recommend that folks try it out, I thought local LLMs would be a pain to setup and use, and ollama made it super easy.
Thanks (to both you and the parent) for sharing these details. So is it fair to say the following:
1. Fine-tuning bakes the knowledge into the model, but getting the "source" of an answer to a specific question becomes cagey and it is unclear if the answer is accurate or just a hallucination.
2. Therefore vector databases, which can provide context to the LLM before it answers, can solve this "citation" problem, BUT:
3. We then have limits because of the context window of the LLM to begin with.
Is that a fair understanding, or have I totally gotten this incorrect?
Edit: Or, are you saying that you both fine-tune AND also use a vector database which stores the embeddings of the dataset used to fine-tune the model?
What's more interesting to me was that the first third or so of the email was just banter/chit-chat about sports and family updates. Good reminder that business is always about relationships, and while cold hard factual analysis (which is the latter part of the email) is important, the social and relationship building elements are key.
I think a big part of it is there are so many "here's how easy it is to set up!!!" guides out there, none of which actually tackle security first, or in a way that reflects the training (or lack thereof) many of these folks who are actually doing the work have.
It's nice to assume that everyone setting up backend services for the multitude of companies out there have gone through accredited training and have years of strong production experience with security chops.
The reality is that because the gap between technically inclined and technically clueless is so large, anyone who can stumble through an online tutorial can be seen as "experienced" to someone who isn't.
I'm not sure what the answer is, but this is going to keep happening - maybe a 3rd party service that evaluates "Getting Started" guides for backend services? If basic security protocols are not covered, they get a red mark, and business owners could use that as an indicator of whether their tech folks could potentially screw it up.
They have real experiences of a union in their company. You have examples of unions in your country. Who is correct?
If your answer is yourself, then you're not understanding the point or the concern that people who are against unions have. It doesn't matter if Sweden implemented it well and somewhere else implemented it poorly. Bottom line is that the poorly-implemented scenario can happen, does happen, and in fact happens more often than anecdata of "oh no true union does that".
Unless you can propose a foolproof way that the "wrong way" is always prevented, then you will get opposition to your beliefs. Calling people or things stupid doesn't solve that.