Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
wsxcde
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
21 ms
·
121.
▲
by
wsxcde
9y ago
Sure, going from recognizing letters to actually reading is what I meant by learning language. If you have some degree of familiarity with a language, you're no longer looking at individual letters and pattern matching them with symbol
122.
▲
by
wsxcde
9y ago
Learning the Latin script isn't particularly hard nor is the script itself that useful in understanding languages like English. Learning scripts, especially the roman script which only has 26 letters takes about a day. Learning a l
123.
▲
by
wsxcde
9y ago
I'm not a radiologist, so I can't answer accurately and I'm not even going to try. If you really want an answer, there's plenty of information on the web. But they're most certainly not just pattern matchers in the
124.
▲
by
wsxcde
9y ago
What do you think roomba is, if not a house-cleaning robot? ED: yes, they're not perfect but give it about 20 years. I'm pretty sure we won't have human cleaners by then.
125.
▲
by
wsxcde
9y ago
Because robots aren't cheap, at least for now. We're still at a stage where compute is cheap but interacting with real world using interfaces that go beyond screens and keyboards is still cumbersome and expensive. But coming back
126.
▲
by
wsxcde
9y ago
If all a radiologist does is pattern matching, the market would've replaced them with techs a long time ago. If not in the US, this would've happened in at least a few other countries.
127.
▲
by
wsxcde
9y ago
You're being biased by presentism. In an age where robots are cheap and ubiquitous, we won't still hire cleaning ladies (interesting assumption they're still ladies) just because they'll be able to clean under the rug. I
128.
▲
by
wsxcde
10y ago
I don't get the hate for overbooking. If you know 3% of passengers won't show up for whatever reason -- they're late, missed connections, cancelled plans -- it makes sense to sell 103% of capacity and give everybody 3% cheape
129.
▲
by
wsxcde
10y ago
As someone who knows a bunch of different scripts (Perso-Arabic/Urdu, Brahmi-based scripts like Devanagari, Kannada, Tamil, Bengali and obviously the Roman script), scripts do make a difference. Yes, a lousy script isn't a fatal i
130.
▲
by
wsxcde
10y ago
> The bidding may go high, but it will only go as high as how much it would cost to hire an American instead. Except this is already the case? Hiring a H1B worker isn't free. It costs something like $10K in legal costs alone. Do you
131.
▲
by
wsxcde
10y ago
It means fixed-point arithmetic. Floating point in SMT doesn't work very well, so that's why they have this restriction.
132.
▲
by
wsxcde
10y ago
STOKE is on github here: https://github.com/StanfordPL/stoke . Most CS conferences are making a big push towards "artifact evaluation," which means that many papers now come with peer-reviewed implementations.
133.
▲
by
wsxcde
10y ago
Checking whether your code is correct under aliasing has nothing to do with the SMT solver. The solver can only analyze the model of the program given to it and it looks like the STOKE memory model assumes that the arguments don't alia
134.
▲
by
wsxcde
10y ago
You still don't get the point. Recognizing the evils of British imperialism is completely orthogonal to disavowing the English language.
135.
▲
by
wsxcde
10y ago
There many languages with multiple scripts. Konkani is written in both Kannada and Devanagari. Manipuri is written in both Manipuri and Bengali. For a more controversial example, I'd say maybe Urdu/Hindi are basically the same lan
136.
▲
by
wsxcde
10y ago
India in its current form never existed, its like the EU This is technically true in that yes, a state with precisely India's borders today did not exist before 1947. But that is a useless claim because that is true for every country
137.
▲
by
wsxcde
10y ago
Yup. There are loads of errors. They're also showing Telugu letters instead of Kannada and Urdu instead of Pashto. Another error I noticed was that the letter (ழ்) in Tamil is shown as [ḻ] like"rag" -- I don't know what
138.
▲
by
wsxcde
10y ago
> most of the owners of these in various national institutes are shopping for customers to justify their funding What's the evidence for this claim? AFAIK the DoE INCITE awards -- which is how you get allocated time on the big sup
139.
▲
by
wsxcde
10y ago
Prefetch instructions are difficult to get right from a compiler perspective. If you prefetch too early, you might kick useful data out of the cache. If you do it too late, it's obviously useless, plus you just wasted an entire instruc
140.
▲
by
wsxcde
11y ago
One more amusing anecdote about "failing" a defence. I was told this happened in the EE department at Princeton. There was a student whose defence wasn't going well and the committee’s consensus was that the more work needed
141.
▲
by
wsxcde
11y ago
Should we also then count the amount of money that are lost to bitcoins being stolen due to malware, poorly written private key generators, exit scams and so forth? I think what I did was a fair comparison.
142.
▲
by
wsxcde
11y ago
No, in terms of the overall cost of operation of the system, bitcoin is most definitely not cheaper. Transaction costs paid by the users are low because of the inflation subsidy paid out to the miners. In terms of overall costs, the bitcoin
143.
▲
by
wsxcde
11y ago
Why is it a monopoly position? All the other SoC makers - Samsung, Apple and AMD - would negotiate similar deals with Netflix.
144.
▲
by
wsxcde
11y ago
Suppose you have a program that uses SGX. Perhaps this program requires some public keys which it uses as a root of trust. Presumably you've baked these public keys into your program, you load this binary with the code+public keys into
145.
▲
by
wsxcde
11y ago
I'm guessing they want to insert themselves into the entertainment content distribution systems and collect some royalties. AFAIK the initial request for SGX came from the paranoid-types at HFT firms who were worried
146.
▲
by
wsxcde
11y ago
> Most people making arguments for "the other side" haven't actually had to defend them against real criticism. If you go into a conversation with a condescending viewpoint like that, don't be surprised if you don
147.
▲
by
wsxcde
11y ago
But VLIW and OoO aren't the only two design points. The renewed interest in traditional inorder vector processors. In any case, I do think the point of DSPs was to be area/power efficient for certain specialized algorithms, and a
148.
▲
by
wsxcde
11y ago
Sure, that's fair. I was just pedantic about the precise meaning of a "synthetic" benchmarks, which is a benchmark generated to mimic real code but isn't a real/useful application by itself.
149.
▲
by
wsxcde
11y ago
I suspect the DSP makers just use VLIW because of interia. They probably don't have the money or incentive to revisit their old decisions. Also wouldn't you say that most of the stuff that used to be implemented on DSPs is now mo
150.
▲
by
wsxcde
11y ago
Patterson says they've tried their best to make sure it doesn't infringe on patents they know about but they obviously can't provide any guarantees.
More ›