sdwr

16 ms·

241.
by sdwr 2y ago
Personal responsibility, not legal liability. In the way a child can be responsible for a pet. Chatgpt was trained on benchmarks and user opinions - "throwing **** at the wall to see what sticks". Responsibility means penalties fo
242.
by sdwr 2y ago
Great improvements and all, but they are still no closer (as of 4o regular) to having a system that can be responsible for work. In math problems, it forgets which variable represents what, in coding questions it invents library fns. I was
243.
by sdwr 2y ago
Yeah, it's friendly ugly. I'd be happy going to work there everyday.
244.
by sdwr 2y ago
The funny thing is that, if anything, we are the dumb ones in this scenario. They figured it out through reasoning and trial and error and raw experience, while for us it's apparently impossible to learn anything we don't already
245.
by sdwr 2y ago
Very true. The definition of realist has flaws...
246.
by sdwr 2y ago
Arguably, every non-standard set is made for aesthetics (and provides a worse playing experience). If you are a chess player, you are thinking strategy and tactics, and the pieces (at best) blend into the mental background, and (at worst) a
247.
by sdwr 2y ago
Idealist: "the free market is perfect!" Realist: "the free market has flaws, which can be addressed by..." Replace "free market" with anything you like
248.
by sdwr 2y ago
There's a constant tendency here to assume other people don't think, they blindly copy what's popular. Everybody learns partially through pattern recognition, partially from trial and error, partially from armchair reasoning.
249.
by sdwr 2y ago
It's a lengthy, private, 1 on 1 interaction with a complete stranger. I've seen nothing but (minor) problems on the uber and lyft subreddits about male drivers and young solo female passengers. More robust processes protect everyo
250.
by sdwr 2y ago
The difference is whether chips with no defects get artificially binned. In a competitive market, if you have a surplus of top-tier chips, you lower prices and make a little more $$ selling more power. With a monopoly (customers are still y
251.
by sdwr 2y ago
The point of the Boltzmann brain is that it's spontaneous (which is also what makes it impossible). If a billion years of slowly ratcheting up complexity counts as a "random fluctuation", then every brain is a Boltzmann brain
252.
by sdwr 2y ago
Plus they get to avoid the pain, chance of partial recovery, messy cleanup for their family, and carrying out a criminal action. Not saying it's right, but there are practical reasons why "dying" looks a lot more attractive t
253.
by sdwr 2y ago
> The psychological hole you can dig yourself with these chatbots is so much deeper than just a typical fandom, especially when you're predisposed to finding it appealing. Spot on. Described pretty even-handedly in the document: &gt
254.
by sdwr 2y ago
Ahh, this is the comment that cleared it up for me. MM takes on risk, can offer tighter spread when not exploited (ex. HFT arbitrage) Could theoretically take advantage by manipulating prices But is already operating within the bounds of th
255.
by sdwr 2y ago
"Because the earth existing is more unlikely than winning the lottery, we should expect to see winning lottery tickets floating in space"
256.
by sdwr 2y ago
Sounds about right, except Pachelbel's Canon actually is the best song ever for non-musicians. It showcases harmony and contrasting lines in the simplest, punchiest, most pleasing way. It's fundamentally "the good stuff"
257.
by sdwr 2y ago
Health insurance relies on young people subsidizing old people. If you lose enough young customers, you end up in a "death spiral" where you can't afford your claims, but if you raise rates more customers will leave. There ar
258.
by sdwr 2y ago
The Unabomber's MO was politically-motivated terror attacks aimed at high-rank targets. That's exactly what the NYC killing was, so it supports the claim that this guy did it.
259.
by sdwr 2y ago
I think the writer here is resisting Kafka's message. In a classic adventure, the hero fights bravely against the odds and succeeds. In Kafka, the hero fights bravely against the odds and fails. Like Camus, Kafka finds vitality (not ho
260.
by sdwr 2y ago
People who value their peace of mind or long term safety might. Reckless, impulsive, emotionally empty people won't be deterred at all
261.
by sdwr 2y ago
If you go through the API (with chatGPT at least), you pay per request and are never limited. I personally hate the feeling of being nickeled-and-dimed, but it might be what you are looking for.
262.
by sdwr 2y ago
Thinking about that is like getting spooked by your reflection in a mirror - instincts being used in a situation where they don't apply. If there's only one you at a time, neither you nor an outside observer can tell the differenc
263.
by sdwr 2y ago
Yeah, you can argue the same thing about going to sleep, there's no guarantee that the same "you" wakes up. From the outside, an "identical clone" is indistinguishable. On the inside, the clone feels exactly how you
264.
by sdwr 2y ago
Saving is disciplined, forward-thinking, and virtuous, in a sober, Puritanical way. Gambling is degenerate, chaotic and immoral. Yotta is a savings account for gamblers. The moral argument is "Gambling is bad, these people tried to bet
265.
by sdwr 2y ago
"the United States is stolen land that should be given away" is a self-disparaging comment - it only works when you're lying. Just like "Oh, don't mind the mess" bridges the tension when your place is basically
266.
by sdwr 2y ago
I've always seen it as a "candle that burns twice as bright lasts half as long" type thing. Like, if you want a car to still be roadworthy in 20 years, how often do you drive it? Not never, not all the time. You want the litt
267.
by sdwr 2y ago
You built a platform to facilitate and track discussions, for people and AI. That's great, and maybe it's the future of harnessing AI, but it's not: - Conversational Game Theory - a dynamic equilibrium between computational,
268.
by sdwr 2y ago
It's saying, "AI needs consistent, scored rules for communication, so it can self-train being a person the same way it self-trains winning games of Go" (sure). Then it says "we have those rules" (they don't) &q
269.
by sdwr 2y ago
Where's the surprise? It's the classic business 2-step - drum up interest with "too good to be true" features, then cut them back. The marginal customers who need those features leave (and are too expensive to keep), eve
270.
by sdwr 2y ago
"Preserving the irreplaceable value of ..." works great as a personal motivation or a mission statement, but terribly as an explicit rule. It's the same problem as carbon offsets - if there's a formula for converting int

More ›