Still experimenting with different ways to make learning easier using LLMs.
I put together Codose as a tool where you paste a link to an Exercism or LeetCode problem, and it spins up a code editor with an AI tutor that walks you through the solution step by step, with mini lessons along the way when you need them.
You can try it without signing up but I’m on the Google AI Studio free tier right now, so I’m not sure how many uses can it handle
Thanks! Is it that you can't see the graph view at all? It uses DeckGL which might be a bit too fancy and may not work on all browsers possibly. It works on Chrome/Safari on my iPhone and Chrome on Arch.
I didn't realize that you actually could provide a working link back to Hacker News but it seems HN does support that. Thanks, I will give that a try!
Oh ok. Yeah, it is just not really complete yet. What this does is displays a UMAP of the embeddings for the selected feature and you can mouse over for labels. It isn't identifying clusters itself yet, finding exemplars, etc. Right now, points that are close to each other are just semantically close - as judged by UMAP.
One thing I’ve always disliked about RSS (and this could actually fix it) is duplicates. When a new LLM model drops for example there are like ~5 blogs about it in my RSS feed saying basically the same thing, and I really only need to read one. Maybe you could collapse similar articles by topic?
Also, would be nice to let users provide a list of feed URLs as a variable instead of hardcoding them.
Looking at the console messages with the LLM reasoning, it does seem to work quite nicely for deduplication. Your example is probably even a lot easier than news articles, where you can have many articles from different viewpoints about the same event.
I don't actually plan to run this as a service so there's some things hard-coded and the setup is a bit difficult as you need an API key and a proxy. Currently it's just experimentation, although if it works well, I'll probably use it personally.
> If you want a half-decent model, at least look at NYC
I don't have the data, honestly, but isn't NYC (and it's surrounding cities/suburbs) more dense than the Bay Area?
In SF (the city) transpiration is quite decent because it's dense; single family houses and public transpiration together is a very very tricky to pull; you have to choose one or the other and most people would rather live in a family house than an apartment/condo with good transpiration
The electric trains are a bit faster, but still slower than something like BART.
My main issues with Caltrain are the cost (public transit shouldn’t be more expensive than driving) and the schedule: depending on the station, missing a train can mean waiting 30-60 minutes.
also the Peninsula isn’t really built around trains. I still have to drive 5-10 minutes just to get to the station, then wait, then ride at that point it’s easier to just drive the whole way. If I lived within walking distance, I might use the train way more.
As long as houses are treated as an investment, there will be all kind of incentives to make them go up in value.
China did "over" solve it by building loads of houses, so everyone's house value just crippled down like crazy.
If buying a house delivers 5–10× the ROI of the S&P 500, that’s not “smart investing” then there is a huge problem to be addressed ... (preferably by building more)
>China did "over" solve it by building loads of houses, so everyone's house value just crippled down like crazy.
That's not really a bad thing unless you believe real estate should be a vehicle for investment. However, in T1 and T2 cities asset prices are roughly on par with western major metro areas. However, property management fees tend to be lower. I suspect the Hukou system contributes significantly to keep housing affordable, and it's not just the massive supply.
5–10× the ROI of the S&P 500 - which is really only true because we allow massive leverage on real estate that we'd not on stocks themselves.
And leverage brings all the problems associated with it; if you own your house outright and it goes down %10 you're out 10%; if you only have 10% equity and the same happens you're down 100%. The gains, of course, work similarly which is why for so long it's been a heads I win, tails I win.
It’s not the healthiest food, but it’s a much weaker risk factor than diets high in processed foods (including processed meats), refined carbs, added sugar, and excess salt.
For adults (25–64), the biggest diet-linked contributors to cardiometabolic death were sugar-sweetened beverages and processed meats. [1]
also form the paper:
High sodium intake → ~66,000 deaths (9.5%)
Low nuts & seeds intake → ~59,000 deaths (8.5%)
High processed meat intake → ~57,000 deaths (8.2%)
Agreed, much of the angst towards "meat" or "red meat" really ought to be stated as carcinogenic nitrosamines . When you cook nitrate rich red meat at high temperatures you get nitrosamines (less when you use herbs like rosemary) ...
Meat has it's own internal ranking: Fish, boneless skinless chicken, low fat red meat cuts, high fat red meat, processed/preserved red meats (like sausage, bacon, sandwich meat etc) ... We should shift left in our meat consumption.
> then just shat all over them with start menus made with React Native, control-alt-delete menus that are actually just webviews, and forcing Copilot down everyone's throats
Thank god I've been using Linux long enough to not experience any of that.
At my job in a large non-tech company, almost everyone uses Windows (except for the dev team) purely because of Microsoft Office. As long as that thing exists, they can do all the dumb things they want and still dominate.
Ironically, MS Office is one of the best working Microsoft software on Linux, through Office 365. It works so well, that on my Windows work computer, I worked for months editing Word docs and Excel sheets every now and then, without realizing I didn't have Ms Office installed.
The funny thing is, MS office apps looks less out-of-place on Linux because most apps are already different. On Windows 11, it looks like the forgotten step-child of Windows 8.
> Dubai has a lower median wage than either London or Mississippi, but people don't think of it as economically broken.
Dubai isn’t sold as a place to belong long-term. Most people move there knowing it’s temporary.
The Bay Area is drifting in the same direction too with the increased cost of living around here. (but the same could be said about most big cities, maybe?)
That's everything at some point, no? Easier to talk about something, harder to start something, and much harder to actually see it through.
Personally, I am making a ray tracer project right now in Rust, so I hope I can become the latter. It being something I did before (albeit, long ago in c++) helps.
Still experimenting with different ways to make learning easier using LLMs.
I put together Codose as a tool where you paste a link to an Exercism or LeetCode problem, and it spins up a code editor with an AI tutor that walks you through the solution step by step, with mini lessons along the way when you need them.
You can try it without signing up but I’m on the Google AI Studio free tier right now, so I’m not sure how many uses can it handle
reply