gfd

Doubtful. Most people (including Mcgee) probably don't want the world to know they didn't pay back a debt for nearly 3 decades

Citrix

If flashcards count as games, then Memrise (or Anki which is FOSS but you have to setup the gamification aspects yourself)

If ELO rating grinding counts, then chess or competitive programming. Or any game with a competitive scene.

Gigabyte gaming monitors come with a physical kvm button: https://www.gigabyte.com/Monitor/M32U

It's in the red inked code review image and it's in his rewrite as "source = source[:-2]"

Isn't `a=a[2:]` going to copy and make that linear time too? I'm surprised he caught the `b.pop(0)` bug but then suggested doing that afterwards.

Also he should submit to codeforces to get a comparable running time. For example here is a random person who submitted AlphaCode's version and it takes 1091 ms https://codeforces.com/contest/1553/submission/144971343

>I start taking English classes at the age of 46

I think it should be "started" instead of "start".

I'm also ESL and have given up on learning English properly after 30 years of trying. Nowadays I just rely on grammarly/gmail (and it does catch this particular example).

I think my only issue with using this is that it forces everyone to use gpt if they want to remain competitive. Which is already the case for many other tools, for example people had to learn z3 smt solvers for previous years.

But the main difference here is that there's no differentiation between the top competitors anymore. It just means in addition to being very good, you still need to run this thing in the background to hedge against this being an easily gpt-able problem while you code up the real solution. And in the cases where it is gptable, you can expect the leaderboard to be ranked only based on latency to openai/aoc servers.

Just a few weeks ago I remember reading Zuckerberg wanting to acquire Unity as part of their leaked VR strategy https://news.ycombinator.com/item?id=33538742

I wonder why they didn't consider unreal instead?

Competitive programming!

20k+ contestants per contest, around 1-2 times per week: https://codeforces.com/contests

In terms of "scene", there are exclusive discord channels that you can only join if you have above a certain rating (usually candidate masters and above). Probably the highest average IQ community that I'm part of and they discuss stuff beyond competitive programming.

iirc someone on hn said the going rate for zero-days is just $1-2 million. there were at least 3 RCE on chrome just this year: https://www.theregister.com/2022/04/15/google-third-fix-chro...

the math works out for someone to spend it here given the amount of panicky crypto whales trying to get their money out of ftx. (OTOH, the math also suggest that these exchanges should've been hacked long long ago if these cheap zero days exist, so idk)

His about page has some of his competitive programming results: https://oimo.io/about

(Sidenote: try dragging the profile pic!)

Only tangentially related, but don't put cash in safety deposit boxes. Police can take them under civil forfeiture laws:

https://nypost.com/2021/06/12/fbi-aims-to-keep-valuables-86m...

https://www.businessinsider.com/fbi-raid-1400-boxes-us-priva...

Not saying it makes it okay, but every company in existence does this to some extent. Everyone does A/B testing with the intent to alter user behavior (aka psychology) to increase their profits.

Was this term coined on HN? I remember first seeing it (used in an AI context) from this 2019 comment under "Cool stuff that's still completely unregulated": https://news.ycombinator.com/item?id=21167689

Most of the predictions in that first comment came true.

Was there a dedicated HN submission for this topic?

This is extremely fascinating to me. In what Andrej Karpathy calls "software 2.0"[1], leaking your model is equivalent to leaking the main IP of your company. Unlike source code where it loses value quickly out of context (e.g. twitch leaked their code yet that didn't spawn a bunch of twitch clones), models can be fine-tuned and transfer-learning-ed for many other purposes

Since these models take millions of dollars to train, I see these sort of hacks becoming a thing! I wonder when companies will start adding "trap streets"[2] to prove that others are using their stolen models?

[1] https://www.youtube.com/watch?v=y57wwucbXR8

[2] https://en.wikipedia.org/wiki/Trap_street

To save someone a click, "sketch" has a technical definition when it comes to data structures (probabilistic data structures like count min sketch https://en.m.wikipedia.org/wiki/Count%E2%80%93min_sketch)

But this link is using the non technical definition (drawings)

I've never bought a bond in my life (I didn't even know what they were) until someone told me they are currently offering 9.62% risk free returns! There's a very low limit on how much you can buy per year but this seems like a no brainer to me.

https://www.treasurydirect.gov/indiv/research/indepth/ibonds...

I was curious what that looked like and turns out there are videos on youtube: https://www.youtube.com/watch?v=60fqruVDrjs

Pretty cool!

Since the honey also happens to be blue in this case, couldn't it be possible that these keepers are wrongfully blaming the M&M's factory? I think the main point of the article was that everyone has their own theory of what the cause is and are all pretty adamant about it:

>Some beekeepers, like Sheppard, swear the berries are to blame for the odd blue honey, which appears intermittently at most, and only in limited hives near the state’s Coastal Plain. Other keepers suspect it might be nectar from the kudzu that grows rampant over nearby fields and forests. Near Dunn, keepers with traces of blue in their hives point to the grapes the bees frequent. And in the eastern part of the state, beekeepers are watchful of the plant Summer Titi, or southern leatherwood.

Why did they close the discussion due to code of conduct? I didn't see anything wrong with the previous comments before that point.

I didn't consider Motorola a Chinese brand but apparently they were sold by Google to Lenovo in 2014 https://en.m.wikipedia.org/wiki/Motorola_Mobility

maybe look into AWS Vault?

I don't recognize the contest name although I did math contests a long time (over ten years) ago.

Why aren't AMC/AIME/USAMO/IMO mentioned?

i remember the last time i saw his name was on https://news.ycombinator.com/item?id=14894653 (which i guess lost against tensorflow.js?) in around 2017 or so.

Did anyone try running this benchmark with other python implementations? (Pypy etc)

The previous big case I remember was linkedin leak with 700M users: https://news.ycombinator.com/item?id=27674393

At this point I've basically accepted that all my info will be found on sites like fastpeoplesearch.com and that anything I tell any company (or I guess in this case, govt too) will eventually be leaked, correlated, and used against me.

“Good, he did not have enough imagination to become a mathematician.

[Upon hearing that one of his students had dropped out to study poetry]”

― David Hilbert

Their founder has a pretty famous blog that frontpages all the time (5+ pages of search results with 100+ upvote each): https://hn.algolia.com/?dateRange=all&page=0&prefix=false&qu...

So no surprise that the product he made is also interesting to HN

I thought this was going to be about https://en.wikipedia.org/wiki/Szemer%C3%A9di_regularity_lemm...