metalspot

10 ms·

91.
by metalspot 3y ago
> US foreign policy apparatchiks who are, unfortunately, pretty much just that. there is an enormous risk in the US, which has not fought a war on its own soil since the civil war, and has not been in a real existential battle since WWII
92.
by metalspot 3y ago
this kind of infantile analysis is about a step above flat earth conspiracy theories at best. it has no place whatsoever in intelligent conversation. a 50 year old bullet may not work. are you going to put a gun in your mouth and test it? (
93.
by metalspot 3y ago
> iterative process: You can simply send a follow up message saying "You did X wrong". It is usually happy to correct itself. cool. i am interested to try it out.
94.
by metalspot 3y ago
there are two factors that control market prices: supply elasticity and demand elasticity. commodities with inelastic supply and inelastic demand are prone to very wide price swings. inelastic supply means that increasing prices don't
95.
by metalspot 3y ago
the concern i would have here is that if you don't know SQL then you have no way of knowing whether or not the results are correct and if the results are wrong you have no way to debug the query. to construct an SQL query you need to k
96.
by metalspot 3y ago
ok, i kind of understand that, but it is still very different than a metric. metrics become problematic because they are reductive and they are tied to compensation, promotion, or other rewards. here we are talking about a raw dump of all d
97.
by metalspot 3y ago
>TikTok and instagram are proof we want this type of content really inspiring take here. this is like saying fentanyl is a great product because people really like it. but i think you are wrong. these apps are mostly used by children and
98.
by metalspot 3y ago
>simplified metrics frequently distort incentives i dont understand how this relates to transparency. nothing about giving patients the right to access their own medical records involves metrics in any way.
99.
by metalspot 3y ago
>Nobody knows how even the current generation of LLMs work we know exactly how they work. all they do is use statistical probabilities with limited randomization to synthesize new outputs from massive samples of inputs. essentially a mac
100.
by metalspot 3y ago
getting a degree is nothing more than passing a series of tests over a period of time. medical licensing is residency, exams, board certifications, etc. it is all tests. the question is who is setting the rules for those tests and whether o
101.
by metalspot 3y ago
> correlates well with job performance highly doubtful, but i would be interested to see your evidence. IQ tests might provide a rough gating mechanism but applying them in the field you don't have any control group so there is no p
102.
by metalspot 3y ago
> companies are generally cautious about this sort of thing yeah, they hire lobbyists to do it for them so they don't have culpability, but you are right.
103.
by metalspot 3y ago
> avoid even the appearance of impropriety the real world standard is don't do anything that knowingly or intentionally violates the law or your employers policies. anyone can falsely claim that what you are doing is wrong but as lo
104.
by metalspot 3y ago
> 500 hours taken consecutively can hardly be compared to 500 hours spent over multiple years 500 consecutive hours should be far superior because the time is focused and with fewer distractions. training people in a professional 9-5 cor
105.
by metalspot 3y ago
> hardly any agriculture is sustainable humans have been doing it for 1000's of years. that is a much longer record of sustainable success than anything else we know of. there isn't any particular reason to think that industria
106.
by metalspot 3y ago
> a private company just isn't going to invest the time and energy needed to train an apprentice not as long as they can force people to pay for it themselves and have the government and the tax payers subsidize it. universities wer
107.
by metalspot 3y ago
> employers had a burden of showing that their tests were sufficiently related to the posted job so why not just create a test that actually tests for the skills needed for the job? that would be the obvious way to optimize profit from a
108.
by metalspot 3y ago
if a job has specific knowledge and skills requirements then it needs to have an objective test that people pass to be able to do the job. that is the only way to eliminate the systemic discrimination that the university system and degree r
109.
by metalspot 3y ago
> GAA was suppose to come in N3 any source on that? as far as i know there was never any official roadmap or announcement by tsmc of gaafet on n3. this article seems to be from the n3 announcement (april edit: 2020) ( https://w
110.
by metalspot 3y ago
you are right. but the claim that tsmc is "2 year behind intel" based on those roadmaps is assuming a lot. intel has consistently failed to deliver on their prior roadmaps and is currently behind tsmc in what they are shipping, so
111.
by metalspot 3y ago
Intel 20A is scheduled for mass production in 2H24 and TSMC N2 in 2H25, so 1 year according to current roadmaps. However, fabs get judged on what they ship, not what the promise. TSMC is shipping N3 in volume, and Intel 4 is still not shipp
112.
by metalspot 3y ago
> smart motivated ones find a way to go to college the bigger sample bias is that wealthy people almost universally send their children to college and their children are always going to earn above average wages. it is pretty obvious that
113.
by metalspot 3y ago
An LLM is still a deterministic machine, which why it is the same as an Abacus. AI today is built using classical computers but the human mind is a quantum computer. AI is a simulation of a quantum computing device built using a classical c
114.
by metalspot 4y ago
> need to rethink how we teach and test students if ChatGPT is good at cheating all that proves is that the curriculum and grading mechanism are garbage, which is already a well known fact, so shouldn't come as any surprise. a schoo
115.
by metalspot 4y ago
> Solar panels in particular are about as dangerous as an inert rock just wait until they are abandoned by bankrupt companies and start to break down and all of the highly toxic substances they are made of leech into the ground water.
116.
by metalspot 4y ago
the nuclear industry has a history of creating the safest power source ever. that is just a fact. as mentioned, everything fossil fuel related has a massively higher death and injury cost than nuclear. the california wildfires caused by the
117.
by metalspot 4y ago
IEEFA is neither independent nor a serious research organization. The fact that they pretend to be only undermines their credibility. The paper you link to is a joke. They claim a cost of utility scale solar as $17 / MWh and then claim
118.
by metalspot 4y ago
> 90% of them are plumbers, and many not very good ones i think this guy understands the industry very well. maybe you are the one with tik-tok opinions?
119.
by metalspot 4y ago
> The primary deciding factor is always security so, anywhere where your workloads or data are physically co-located on the same hardware as someone else's should be automatically disqualified, right?
120.
by metalspot 4y ago
not credible ones. google "ex post facto law" if you are confused about this.

More ›