StevenWaterman

6 ms·

61.
by StevenWaterman 1y ago
I'm glad I'm not the only one thinking that. These are such minutiae. Where's the discussion about humans? They're probably the most important part of your system, and the most chaotic, and the part that needs the most c
62.
by StevenWaterman 1y ago
Essentially the idea is that there is an "optimal" amount of alertness (inverted U curve). People with ADHD start below the optimal point, and stimulants move them up towards the optimal point. People without ADHD are typically cl
63.
by StevenWaterman 1y ago
We already have a phrase for "if you start a fight you get punished" The phrase is "normal, non-zero-tolerance policy"
64.
by StevenWaterman 1y ago
https://en.wikipedia.org/wiki/Zero_tolerance > A zero-tolerance policy is one which imposes a punishment for every infraction of a stated rule. Zero-tolerance policies forbid people in positions of authority from ex
65.
by StevenWaterman 1y ago
You see this in ADHD groups where someone will start stimulant medication for the first time and say "This is incredible. I can't believe this is what everyone else feels like all the time" And the crowd emerges to reinforce
66.
by StevenWaterman 1y ago
> It would be weird if stimulants only had an effect of ADHD patients One example of this actually happening is the concept of a "stimulant nap" in people with ADHD, where stimulants actually make them sleepier. Also manifestin
67.
by StevenWaterman 1y ago
What do you call system design, when it's referring to the design of systems in general, and not just computer services? As in: - writing a constitution - designing API for good DX - improving corporate culture I intuitively want to ca
68.
by StevenWaterman 1y ago
And likewise, a single neuron is clearly not conscious. I'm increasingly convinced that intelligence (and maybe some form of consciousness?) is an emergent property of sufficiently-large systems. But that's a can of worms. Is an a
69.
by StevenWaterman 1y ago
Yeah, I think this is exactly it. If something sounds extremely stupid, or if everyone around you says it's extremely stupid, it probably is. If you can't justify it, it's probably because you have failed to find the reason i
70.
by StevenWaterman 1y ago
That paper was that thinking more makes models worse at easy problems, not hard ones
71.
by StevenWaterman 1y ago
I might have to adopt this way of differeniating OR and XOR in English https://ygdp.yale.edu/phenomena/alternative-one
72.
by StevenWaterman 1y ago
I definitely agree with the difference in formality. > You should try and recalibrate your photon detector I can totally imagine this, in a lab where all the equipment is old, and out of calibration, and the person saying it knows there
73.
by StevenWaterman 1y ago
I'd describe it as: - "try and" implies that the reason for failure is slightly more likely to be from laziness / not actually attempting it - "try to" implies that the reason for failure is slightly more likel
74.
by StevenWaterman 1y ago
I disagree that increased employment and increased labour always makes the pie bigger. If minimum wage was low enough, we would decommission our cement mixers and use a human with a shovel instead. But that's not an improvement. Automa
75.
by StevenWaterman 1y ago
The total salary would go down if you did that
76.
by StevenWaterman 1y ago
If the total salary has gone up, for less work done, it is a positive change. You can solve the inequal distribution via taxes and benefits. Start: 100 people paid $100 After minimum wage change: 90 people paid $125, 10 people paid $0 After
77.
by StevenWaterman 1y ago
Pineapple has a +22 approval rating https://d3nkl3psvxxpe9.cloudfront.net/documents/Pizza_poll_r...
78.
by StevenWaterman 1y ago
The AI's idea of developing a startup is eerily reminiscent of a hacking scene in CSI
79.
by StevenWaterman 1y ago
> Norwegians are taking much more sick leave than a decade ago, driving up costs for health services. Student test scores have worsened more than in other Scandinavian countries ...COVID? Norwegians are taking more sick leave after the l
80.
by StevenWaterman 1y ago
Your question reminds me of the image showing how SpaceX raptor motors evolved https://imgur.com/a/4w3q3lS
81.
by StevenWaterman 1y ago
Yep, every AI call is essentially just asking it to predict what the next word is after: <system> You are a helpful assistant. </system> <user> Why is the sky blue? </user> <assistant>
82.
by StevenWaterman 1y ago
"Weights available" perhaps
83.
by StevenWaterman 1y ago
Are you getting confused between tariffs and VAT? VAT is the equivalent of sales tax. From what I can tell, the tariff on "Potatoes, Thin slices, fried or baked, whether or not salted or flavoured, in airtight packings, suitable for im
84.
Authenticating JavaScript WebSockets(stevenwaterman.uk)
3 points by StevenWaterman 1y ago|0 comments
85.
by StevenWaterman 1y ago
Both "has" and "have" sound ok to me here as a native speaker. Both of these can describe the same event: - The cows have 10 seconds to enter the field - The herd has 10 seconds to enter the field In the case of "cr
86.
by StevenWaterman 1y ago
No, even if the benchmarks are private, it's still an issue. Because you can overfit to the benchmark by trying X random variations of the model, and picking the one that performs best on the benchmark It's similar to how I can pa
87.
by StevenWaterman 1y ago
The paper doesn't seem to actually define their categories anywhere. Am I just missing it? Or am I meant to intuitively know the difference between "spied on / watched" and "spy / surveillance"? Edit - upo
88.
by StevenWaterman 2y ago
She is a cis woman.
89.
by StevenWaterman 2y ago
The noun phrase was "legal defence team" not "legal defence". A decent team can put forward a poor defence
90.
by StevenWaterman 2y ago
On the road, you have to go the speed limit whether you're in a rolls or a lambo. You might prefer a lambo if it could get there twice as fast

More ›