theodorewiles

why dont they increase class sizes

If I had to guess it sounds like they are using CURE to cluster the source documents, then map each generated fact back to the best-matching cluster, and finally test whether the best-matching cluster actually provides / supports the fact?

yes i have been thinking about this for some time. One other thing with zettle I'm not sure was implemented here is you can have topic notes that just refer / summarize other notes; it would be very interesting if these could be autonomously created by some kind of clustering algorithm based on underlying links. Kind of like summary-of-summary.

Also curious if there might be some improvements if you dont rely on semantic similarity and just do all the pairwise "how related are these memories and in what way" LLM test like https://www.superagent.sh/blog/reag-reasoning-augmented-gene....

I honestly don't know why this isn't a bigger deal. This is terrifying.

welcome to parenting

I don't want to throw yet more GenAI crap at this, but could RAG be a possible innovation that unlocks microtransactions?

I'm imagining a backend database of creator-submitted content. The LLM runs RAG on it, pays the creators it relied on to synthesize answers a microtransaction. Then also sells the Q-and-A as a subscription service to the end user. Maybe payouts are conditional on positive user feedback. The solution can also flag queries it sees that don't have great answers yet.

Same moat as search, not reliant on advertising.

yes this is all of a sudden super relevant with LLMs. the challenge of lifelogging was search and summarization.

Can anyone suggest better materials / guides for early literacy development?

i listened to some paganini based on a twoset video. blew my mind - dude was a rock star.

thanks!

FYI you can get dewey decimals programmatically by hacking the OCLC API: http://classify.oclc.org/classify2/api_docs/classify.html

this says you need an API key but I think I found some way to call this without one… You might be able to improve classification by either incorporating the dewey decimals of the books mentioned on a podcast or fine-tuning a model based on known book titles (or maybe there are book summaries somewhere) to known dewey decimals from OCLC.

I emailed you but did something similar and posted to Show HN a while back -

https://weekend-collection.s3.amazonaws.com/Catalog+-+Feb+17...

I SLAMMED coffee the first few weeks and would also drink like a gallon of water a day. Helped a lot. I think the key is to choose the schedule that works for you.

I’ve lost 20 lbs on IF and this is spot on. It’s just a lot easier (for me) to eat way less. I just only eat after 5pm every day. Much much less for me to think about vs any other diet. Its tough at first but mind and body adapted after a few weeks.

Not sure I agree with the real estate conclusion in this market. I think there is a lot of uncertainty on how real estate values will change post-COVID as how we use places for shopping, working, and travel change (or dont)

Just a friendly reminder that the current communist regime in mainland China is not legitimate - the republic of China (currently in Taiwan) is the original member of both the League of Nations and the UN, and has actually transitioned to a stable, transparent, and prosperous democracy, and should be in charge of all of China.

We know:

https://www.bloomberg.com/opinion/articles/2021-02-17/citi-c...

New Yorker poetry podcast worked for me - great explanation from smart poets about what they like about contemporary poems.

https://www.newyorker.com/podcast/poetry

This is a good article.

Instead of jumping to a recommendation engine, I would think about the analog solution (e.g. nybooks.com, newcriterion.com).

Pretty much has solved my book recommendation pain point.

I think they should have to disclose the price they sold them for - would be a great way to get a real valuation on the degree.

Wouldn’t the least abandoned books be more useful?

Yep it reminds me a lot of the fundamental message of Spin Selling - the best salespeople are the ones who figure out the real needs of their customers.

My pet theory is that McK is re-orging the NYT to be more commercial and the writing staff is rebelling, while Jay Alix is more than willing to help out.

Metafilter

I would also look into PE corporate transaction work. This is a growing market (PE deal volume is up), I’m assuming most offering memorandum, etc. are very boilerplate, and I’m also assuming the clients would appreciate the differentiated up-front pricing model. Likely to be repeat customers as they could do a few deals per year, and the community is tight-knit so WOM would work. Finally, you already probably know this but I am assuming there is some overlap between the VC investment community and the PE investment community so you probably already have accounts today that could be natural entry points / “customer zero” for the adjacency. Just my 2c. Love thinking about this stuff.

thanks!

I believe it's ignored

Original author here - although a lot of changes have been merged in! Glad to see people like this one.

I think the insight is that you don’t really have to track down the points for everything you eat. you can guesstimate how many points something is and the measurement error will come out in the wash after you calibrate it to observes weight loss / gain. Pretty neat IMHO!