jnwatson

9 ms·

91.
by jnwatson 4mo ago
Probably just the ol' HN hug of death. They load fine now.
92.
by jnwatson 4mo ago
The solution is obviously the American one. Make everyone so afraid of their job prospects that working for a startup isn't materially different in terms of job security.
93.
by jnwatson 4mo ago
But they aren't DAGs. They are multidigraphs.
94.
by jnwatson 4mo ago
The general idea is that formal methods are self-verifying, up to a point. Sloppy formal methods simply won't prove. The point up to which formal methods stops is: do the formally encoded requirements actually encode what I want to be
95.
by jnwatson 4mo ago
The only measures we see are the surface-level ones, because those are the only ones that sort of work. Alignment is a hard, possibly impossible problem. Anthropic's gambit is they luck upon a solution before the paperclip maximizers
96.
by jnwatson 4mo ago
Correct.
97.
by jnwatson 4mo ago
Back in the elden days, I took a course called "Large Scale Scientific Computing". It was mostly about multiplying large matrices. I didn't think this was going to be remotely applicable to anything commercial. Boy was I wr
98.
by jnwatson 4mo ago
Taken honest, we don't have a large enough sample size to realistically say that humans behave all that differently. There have only been a handful of conflicts where tactical nukes realistically were on the table. Famously, General Ma
99.
by jnwatson 4mo ago
It is far ethically superior for Stanford grads to stick to their wheelhouse: optimizing engagement so that their users doomscroll longer.
100.
by jnwatson 4mo ago
In Apple's press release[1], they mention proposing something similar to what you're talking about. "Given the serious risks to users, Apple designed a solution called Trusted System Agent — an intermediary that would allow v
101.
by jnwatson 4mo ago
If you need a quick patch, you pass another parameter to turn off the 1 day. 1 day delay will prevent more problems than it makes.
102.
by jnwatson 4mo ago
It is going to be awful confusing if Apple starts using Mach the language.
103.
by jnwatson 4mo ago
Geneva is the 2nd most expensive city to live in (behind Zurich). I commend any attempt at moderating prices by building more housing.
104.
by jnwatson 4mo ago
Google can't track information in private inference. That's kinda the point.
105.
by jnwatson 4mo ago
Among the oldest value models, the Dividend Discount Model, says that the value of a company's stock is based on the present value of its future dividend payments. Even if a company doesn't currently pay dividends, it will eventua
106.
by jnwatson 4mo ago
And Germany's official name is "Bundesrepublik Deutschland". I get to call it Germany though.
107.
by jnwatson 4mo ago
I had to remind myself what Haiku is even for. Anthropic hasn't spent a lot of recent marketing on it. When I need a light model, I reach for Sonnet. It is nearly free on the max plans, and quite fast. I don't see a place for Hai
108.
by jnwatson 5mo ago
ETFs are supposed to hold the underlying asset, just like mutual funds. In the worst case, you are due your share of the underlying asset. ETNs (which are sometimes lumped with ETFs but are actually distinct products) have counterparty ris
109.
by jnwatson 5mo ago
You're holding it (Python) wrong. Python OO was a counter reaction to the bondage and discipline that languages like C++ had with private members and protected inheritance. If you have members that users probably shouldn't touch,
110.
by jnwatson 5mo ago
I stopped playing MtG around 1994. It is extremely unusual to see decks of cards I mostly recognize.
111.
by jnwatson 5mo ago
I lost my sense of smell from a minor virus I caught a couple years ago. It probably wasn't COVID (I tested negative at least). It came back very slowly, and unevenly. My coffee/chocolate taste is still quite dim. Of all the poss
112.
by jnwatson 5mo ago
Yeah. The same processes that allow corporations to outsource their software to barely qualified 3rd-world body shops are the processes that allow you to deploy AI-generated code of unknown quality.
113.
by jnwatson 5mo ago
"Port of large memory unsafe codebase has a memory safety bug, news at 11." I don't see what the big deal is here.
114.
by jnwatson 5mo ago
Indeed this was caught by a well-known tool, Miri, that detected this error.
115.
by jnwatson 5mo ago
Randomizing the kernel location is of marginal utility at best. There are so many info leaks that KASLR ends up being only a small speed bump on the way to exploitation. Here's a cool project that inventories all your KASLR info leaks:
116.
by jnwatson 5mo ago
The rule used to be if you used more than two JOIN clauses, it is AI.
117.
by jnwatson 5mo ago
It just struck me. I always thought I had writing software to fall back on, in case my main gig doesn't work out. I don't think it will still be there when I'm ready to return.
118.
by jnwatson 5mo ago
You have your own read cache then? You're nearly at the Oracle-level of writing your own filesystem at this point.
119.
by jnwatson 5mo ago
If you're bypassing the page cache, what invalidates the page cache so that the next read (from the filesystem) isn't stale?
120.
by jnwatson 5mo ago
POSIX is backwards in so many ways. It freezes in time the OS innovations of the 70s.

More ›