m-i-l

11 ms·

91.
by m-i-l 3y ago
https://michael-lewis.com/ Infrequent but longer articles on various subjects (topics at https://michael-lewis.com/categories/ ). See also the blog search at https://searchmysite.net/
92.
by m-i-l 3y ago
Note that the author is saying that he couldn't find any proof that the story is true (e.g. a primary source or corroborating evidence) from his internet searches (Google, Hacker News etc.) But just because something is not on the inte
93.
by m-i-l 3y ago
> "Small data-sets were definitely a limiting factor as well as limited computer power." Not just small data-sets and limited computer power, but also very few libraries to help you out - although you could download something
94.
by m-i-l 3y ago
I went without a car for decades, even in the first few years of having children, and reckon I saved >GBP200K as a result (plus kept in better shape etc.). But that was with the benefit of: - living in cities with good public transport (
95.
by m-i-l 3y ago
Note that Neeva wasn't exactly a small player, managing to burn through $77.5M in VC funding over the space of just 3 years[0]. Note also that, at least at the start of that 3 year period, they simply bought in results from Bing rather
96.
by m-i-l 3y ago
> "Nice, but somehow it feels like I've heard these songs before. Could that be possible?" According to the paper[0] "We found that only a tiny fraction of examples was memorized exactly, while for 1% of the examples
97.
by m-i-l 3y ago
> I'm surprised culturing is reduced to a single throwaway line Yes, as a (lapsed) amateur butter maker, I'd say this is the main difference between European butter and US butter, not the fat content. If you make your own but
98.
by m-i-l 4y ago
> "Dam’s company ... had filed for a US design copyright back in 1961. And in 1965, it took action in court against a company that it claimed was infringing on its trolls without proper license... The court found that Dam Things ha
99.
by m-i-l 4y ago
There are a couple of challenges with the word vectorization + cosine distance approach: 1. The current embedding models are for relatively small amounts of content rather than whole web pages. For example, most of the popular Huggingface s
100.
by m-i-l 4y ago
I clicked on the article hoping to find information on potential new cognitive skills humans will need to learn to differentiate Large Language Model (LLM) hallucinated facts from real facts, but unfortunately the article doesn't touch
101.
by m-i-l 4y ago
I always try to maximise postal information rather than minimise, to reduce chance of error, given in my case I have many opportunities for people to make mistakes, e.g.: - I have someone with the same first name and same last name at the s
102.
by m-i-l 4y ago
It was a fast moving corporate production line with a very long queue behind, not a bespoke photo session (I'm not high enough up for that) - less like an agricultural show, more like an abattoir. And the 2 page waiver made me doubt it
103.
by m-i-l 4y ago
Just did one of these last week. First issue was that I am quite tall and the photographer was quite short, but he didn't have a step (presumably for health and safety reasons), so to get the angle right he asked me to squat a little.
104.
by m-i-l 4y ago
In answer to the question "Do you know the location of the Sequoia (Redwood) tree?" there are two - see details and photos at http://www.redwoodworld.co.uk/picturepages/hampstead.htm . Worth exploring hampst
105.
by m-i-l 4y ago
> It feels like the Internet is a more hostile place than ever for small-time websites. You get squeezed from below by wily criminals, and crushed from above by careless megacorps who want to filter out anything that doesn't make t
106.
by m-i-l 4y ago
See also /now pages[0] (a federated version of your gist points 1 and 2, i.e. creating and updating a post about what you are doing now)? [0] https://nownownow.com/about
107.
by m-i-l 4y ago
> "Haven't they all signed the Giving Pledge?" While choosing to give most of your money to various charities, generally of your choosing and while you are still alive to see the benefits, is all well and good, it is ver
108.
by m-i-l 4y ago
> "In 100 years the descendants of Bezos won't be the ruling class in the same way the Vanderbilts, Astors, Rockefellers, Carnegies and Morgans aren't." Andrew Carnegie believed in 100% inheritance tax, on the condi
109.
by m-i-l 4y ago
Random trivia I learned from a visit to a fruit farm a few weeks back: Fruit farmers much prefer bumble bees to honey bees because they're much better fruit tree pollinators - bumble bees start much earlier in the day than honey bees (
110.
UK to become 'true superpower' with first rocket launch(news.sky.com)
1 points by m-i-l 4y ago|0 comments
111.
by m-i-l 4y ago
Focus on the things that you can control and change, and don't fret about the things you can't.
112.
by m-i-l 4y ago
> "Maps is how I usually search for e.g. plumbers." I used to search for tradespeople in this way, but then I realised there's a big industry of while-labelled SEO-friendly companies which register 100s or 1000s of names
113.
by m-i-l 4y ago
All the AI algorithms at the moment need to be trained with human input, e.g. Copilot works on largely human written code, GPT-3 is 60% Common Crawl[0], DALL-E uses 400 million images scraped from the internet[1], etc. I suspect weird thing
114.
by m-i-l 4y ago
I know there's a whole genre (steampunk) dedicated to "what if the Difference Engine had actually been built in Babbage's lifetime and started a steam based computer revolution a century earlier than the electrical computer r
115.
by m-i-l 4y ago
The general purpose search engines search the whole internet, and as a result claim that you can search for anything on the whole internet, even going beyond that to answer questions which aren't on the internet as such, e.g. "Wha
116.
by m-i-l 4y ago
Thanks V. I'm seeing a similar number of problem search requests (although nowhere near as many real search requests:-), so it is probably the same "SEO practitioners" running the same "scraping footprints" against
117.
by m-i-l 4y ago
Thanks for wiby.me. I have seen QHub coming up in the scraping footprints, but my assumption has been that the footprint query is looking for Question and Answers sites powered by QHub containing their targeted terms, e.g. because there&#x2
118.
by m-i-l 4y ago
Actually seirdy.one added searchmysite.net to his excellent list[0] way back in March 2021[1]. [0] https://seirdy.one/2021/03/10/search-engines-with-own-indexe... [1] https://git.sr.ht/~seirdy
119.
by m-i-l 4y ago
Hi, "this guy" here:-) If people come to a site but don't come back then it is reasonable to conclude that "it's a bad website", but as the blog entry put it "without any real users in the first place it i
120.
by m-i-l 4y ago
Thanks for your feedback. It is just the home page which is moderated before indexing (and reviewed annually). When https://personalsit.es/ was listed it looked legitimate, but agreed the results for that site look infected

More ›