HN user

dash2

8,363 karma

Social scientist, hack the occasional R package. My substack: https://wyclif.substack.com. I wrote a book: https://www.wyclifsdust.com. davidhughjones at gmail

Posts145
Comments2,452
View on HN
www.siliconcontinent.com 14d ago

How to avoid being held up by the labs

dash2
2pts0
intheweights.com 16d ago

In the Weights

dash2
2pts0
academic-freedom-index.net 1mo ago

Academic freedom declined in 50 countries

dash2
2pts0
jmlr.csail.mit.edu 1mo ago

A tutorial on conformal prediction [pdf]

dash2
3pts0
hughjonesd.github.io 2mo ago

Mypaintr: Plot R graphics like a human

dash2
3pts0
pickyglutton.com 4mo ago

I had a stroke in Nepal and this is what I ate while I recovered

dash2
2pts0
masslibraryofthings.netlify.app 5mo ago

Library of Things – tools, technology, and equipment available to borrow

dash2
3pts0
hughjonesd.github.io 6mo ago

A slow app to teach the alphabet

dash2
1pts2
news.ycombinator.com 7mo ago

Are you vibe-coding an open source project?

dash2
2pts0
www.trelawnysarmy.org 7mo ago

The Song of the Western Men

dash2
2pts0
hughjonesd.github.io 8mo ago

Why I Wrote Rmlx

dash2
2pts0
hughjonesd.github.io 8mo ago

R interface to Apple's MLX library

dash2
22pts2
wyclif.substack.com 9mo ago

Training AI will take longer than you think

dash2
2pts0
sfshortstories.com 9mo ago

The Answer (1954)

dash2
37pts23
rust-trends.com 10mo ago

Rust's Enterprise Breakthrough Year

dash2
3pts0
wwlln.net 12mo ago

World Wide Lightning Network

dash2
3pts0
www.wired.com 1y ago

Is That Painting a Lost Masterpiece or a Fraud? Let's Ask AI

dash2
2pts0
www.eatingpolicy.com 1y ago

Jed

dash2
2pts0
substack.com 1y ago

"Draw me a 3 panel cartoon "

dash2
3pts1
wyclif.substack.com 1y ago

Cultural Endosymbionts

dash2
3pts0
wyclif.substack.com 1y ago

Makers and takers: why wars are not always won by the wealthiest country

dash2
2pts0
18theses.com 1y ago

The Defense Reformation

dash2
1pts0
twitter.com 1y ago

Draw a Crocodile Without a Tail

dash2
25pts12
arxiv.org 1y ago

ForecastBench: A Dynamic Benchmark of AI Forecasting Capabilities

dash2
2pts0
www.bbc.co.uk 1y ago

Several arrested after woman dies in "suicide pod"

dash2
4pts0
polmeth.org 1y ago

"Shouldn't that work against me?"

dash2
2pts0
assets-eu.researchsquare.com 2y ago

Homomorphic Encryption: An Application to Polygenic Risk Scores [pdf]

dash2
1pts0
www.justice.gov 2y ago

Professor Charged for Operating Multimillion-Dollar Grant Fraud Scheme

dash2
7pts0
hughjonesd.github.io 2y ago

What packages belong together? Learning from R code samples

dash2
1pts0
wyclif.substack.com 2y ago

The academic literature on smacking children is not load-bearing

dash2
1pts0

In Amodei’s Dwarkesh podcast he says that they are constantly estimating the increase in demand for the next leg up, then going out to raise money to build for it. So it’s not necessarily that they are raising money because they are unprofitable.

Checking out a parallel copy of our Go repository and telling the AI to rewrite the whole thing in Zig while I work on something else just so I can keep my job.

Was it just sales fluff? The answer was a lot more interesting.... Executives at their customers were saying absurd things about achieving 100x productivity, and this meant that if any executive at the vendor said that these gains were not plausible, it would undermine the credibility of the customer’s executive, be perceived as an attack (or heresy), and possibly result in an enterprise contract cancellation.

A lot of excellent anecdotes here.

And the moment that happens, Taiwan will instantly vote to rejoin china.

Sounds inflammatory, and also extremely unlikely and has no evidence for it, and also the hyperlinked article, that is supposed to back it up, absolutely does not even try to do so. I don't think reading further is worthwhile.

If you can't sell clothes to anyone, then it may be more socially efficient to destroy them than (a) keep them in a warehouse or (b) ship them overseas. Both (a) and (b) can have substantial environmental costs. I don't think it should be hard to come up with other plausible cases. You're assuming there's only one reason companies do this. I don't deny that that's a possible reason. I also don't see why taxing that behaviour would not reduce it.

How do we know that Chinese models would not do the same? What makes you so sure that China is less likely to abuse my data?

Honour is not the relevant point. What's relevant is that breaking your TOS would give many large individual companies (a) a clear case to sue you and make a lot of money (b) a reason to avoid using you ever again, because who wants their data leaking into a potential competitor? Scraping the internet, as well as being much harder to adjudicate legally, also creates far fewer powerful companies with means, motive and opportunity to take you to court.

A game rather like this is described in Gwen Raverat’s Edwardian memoir, Period Piece. There you have to keep finding new words one letter longer, and they get ridiculous combos like TRUMPET, STRUMPET, (thou) TRUMPETST, STRUMPETST…