foooorsyth

8 ms·

91.
by foooorsyth 3y ago
>Expansion greatly dilutes the fixed costs Where is the actual evidence of this, though? I'm not trying to be snide -- genuinely curious. Waymo is expanding -- their service areas (and operating times) have expanded in SF and Phoe
92.
by foooorsyth 3y ago
>There's literally no choice but to sign the things - regardless of terms - if you want your users to have software that runs. This gets to be a real nightmare in large organizations with multiple Apple Dev Portal admins, some of wh
93.
by foooorsyth 3y ago
> and that regime is more than enough to make a profitable taxi service Waymo isn’t profitable and has been a massive money loser over the past several years. See “Other Bets” in Alphabet’s earnings. Where does this claim come from?
94.
by foooorsyth 3y ago
Software has to be maintained. Nobody wants to do maintenance of their own works — it doesn’t look like a new accomplishment in a performance cycle. Nobody wants to inherit maintenance work. Google’s perf cycle directly causes its graveyard
95.
by foooorsyth 3y ago
Doing it before the holidays is bad PR
96.
by foooorsyth 3y ago
Some VCs will withhold investment until you use a product like Carta or Pulley. Cap table in Excel carries major risk. (And yes, Carta/Pulley are often portcos of those VCs pushing you to use them)
97.
by foooorsyth 3y ago
Limiting speed (below highway speeds) can be incredibly dangerous. Not being able to merge at speed is a non starter.
98.
by foooorsyth 3y ago
Dead man’s (safety) switch
99.
by foooorsyth 3y ago
>I support being paid for your work, but live reload is a fairly trivial thing to write for yourself -- I spent one day this gone weekend writing a debugger in C which does it. Major Dropbox “who would pay for this over just using rsync?
100.
by foooorsyth 3y ago
> My father used to do this too in Boston, but instead by jamming the slot with some kind of paper past the point of removing it This is why the Dunkin flip up lid was such a technological innovation: my dad was worried that he’d get rep
101.
by foooorsyth 3y ago
In Boston in the '90s and early 2000s, an out of order parking meter meant you could park for free. There was a sensor in the coin slot to detect jams which would in turn put the meter out of order. If you ripped off the flip up lid fr
102.
by foooorsyth 3y ago
>Saying that Germany and Japan were “bombed back into the Stone Age” is a wild exaggeration Japan was firebombed pretty ruthlessly. McNamara openly stated that he would have been tried as a war criminal had the US lost the war. Here is a
103.
by foooorsyth 3y ago
> e.g. quotes "", do not do what you think in Google So they changed this behavior without telling anyone then? Because before Google search quality went down the drain, quotes meant to search for an exact phrase. If they chang
104.
by foooorsyth 3y ago
>At that point, all Beeper becomes is a service to turn a green bubble blue. This is a bigger deal than encryption itself to most people using Beeper. Beeper requires you to divulge your Apple ID password to Beeper anyways, so arguing fo
105.
by foooorsyth 3y ago
>Any ideas for how to redesign these? The thought is compelling. I recommend anyone interested in the reshaping of EVs for the future take a look at Volvo's "Future of design" video from a couple years back: https:/&
106.
by foooorsyth 3y ago
>GPT4 is a proof of concept more than anything Hilarious walk-back. “Text can generalize anything” —-> “It’s just a demo, bro” in the same post. Lmao
107.
by foooorsyth 3y ago
>GPT-4 was trained on text And it shows. It has a poor grasp of reality. It does a poor job with complex tasks. It cannot be trusted with specialized tasks typically done by expert humans. It is certainly an amazing technical achievement
108.
by foooorsyth 3y ago
Text and 2D images are a tiny subset of physical reality as perceived by an able-bodied human. Even our best approximation (3D VR headset with Spatial Audio) is a poor representation. We don’t even bother to simulate touch, temperature, equ
109.
by foooorsyth 3y ago
The training space is more important. I don’t think a general intelligence will spawn from text corpuses. A person only able to consume text to learn would be considered severely disabled. A significant part of intelligence comes from exist
110.
by foooorsyth 3y ago
Hanlon's razor, basically. The most plausible scenario here is that the board is comprised of people lacking in foresight who did something stupid. A lot of people are generating a 5D chess plot orchestrated by Microsoft in their heads
111.
by foooorsyth 3y ago
> This is ML, not Software engineering. Money wins, not engineering. Same as it with Google, which won because they invested massively into edge nodes, winning the ping race (fastest results), not the best results. This is an absurd retc
112.
by foooorsyth 3y ago
The guardrails they put on it to prevent it from saying something controversial (from the perspective of the political climate of modern day San Francisco) make the model far less effective that it could be.
113.
by foooorsyth 3y ago
I loved my cube at my first job. We even had a biometric door that separated engineering from the rest of the company. Only engineers behind the door. Juniors and mid levels working in cubes. A few tenured engineers, the primary engineering
114.
by foooorsyth 3y ago
That’s exactly how it should be. Without nation to nation competition, taxes would be even more outrageous than they already are. The institution that monopolizes violence in order to tax should be at the very least threatened by the risk o
115.
by foooorsyth 3y ago
>It would be nice to know what points of friction exist. Is this a serious question? You can't buy food at the grocery store with crypto. Employees likely don't even know how to accept a salary in crypto. Your landlord doesn&#x
116.
by foooorsyth 3y ago
Do you think every acute reaction in the body to intense exercise has to make "evolutionary sense"? Do shin splints make evolutionary sense?
117.
by foooorsyth 3y ago
The context of this exact discussion is the defense of the ugly Rust code above by Rust aficionados who see nothing wrong with it. That code came from the Embedded Rust working group in their learning material for Rust beginners. I'm n
118.
by foooorsyth 3y ago
lol POSIX threads are far more battle-proven than any threading API in Rust. All of the underlying multi-threading on the system you made your comment from is using pthreads. It works fine in reality, despite all the pearl clutching. >Ru
119.
by foooorsyth 3y ago
>it's worse in C pthread_mutex_lock(&mutex); setBit(&gpioCtx, bitPosition, 1); pthread_mutex_unlock(&mutex); It's not.
120.
by foooorsyth 3y ago
>You can also write it more simply in Rust But this is never done, because the language attracts people who overengineer. The original commenter was following a beginner’s tutorial in Embedded Rust. They didn’t just unluckily stumble upo

More ›