dnautics

3 ms·

61.
by dnautics 1mo ago
As a former academic scientist, I suspect it's not gonna work out like they think it will--the process will self select for mostly lemons.
62.
by dnautics 1mo ago
So you're technically correct, but IIUC Tesla is still the most delivered (customers taking delivery) EV brand in China. Draw what conclusions you will from that fun fact
63.
by dnautics 1mo ago
is it? They extended the study from "just wales" to "all of UK" and I wonder if it's possible that in the rest of the UK, adherence to the strict week-of-eligibilty criteria was less strict than in wales.
64.
by dnautics 1mo ago
isnt its core literally femtolisp, ehich was written by one of the julia authors?
65.
by dnautics 1mo ago
by my understanding: there is an extensive and complicated cooling system that permeates the wafer. some of the cores are completely turned off because they fail qc (see the tsmc logo) if all of their neighbors have been going at full bore
66.
by dnautics 1mo ago
That's not really right either. think about it this way: what would happen if the budget was balanced and the us suddenly had a surplus? you HAVE to print money, where is the interest on debt going to come from.
67.
by dnautics 1mo ago
needs an sla that says power will never ever ever go out or else you will have a useless shattered plate of silicon.
68.
by dnautics 1mo ago
this is just a regular reminder that this is by design . https://archive.nytimes.com/krugman.blogs.nytimes.com/2010/0... "in the long run, its really really hard to cut nominal wages. [without] inflation, a
69.
by dnautics 1mo ago
everyone can cherry pick data to whatever economic conclusion they seek
70.
by dnautics 1mo ago
i dont get it, they are comparing eOS to graphene, not talking about the hardware. this would be like grapheneOS not using pixel because the stock android that ships with it does not respect privacy. am i missing something?
71.
by dnautics 1mo ago
calyx os seems serious?
72.
by dnautics 1mo ago
Government funding is political full stop. it might be bipartisan but it's still political, by definition. if you cannot accept this your policy opinions will always be hopelessly distorted.
73.
by dnautics 1mo ago
> There simply isn't enough of a known payoff for private investors to bother and private philanthropists have never come close to making up the shortfall. not true. in the us, the government funds only about half of all research a
74.
by dnautics 1mo ago
btw i have a phd in the hard sciences so i know what im talking about when i say grants are "barely earned privileges". they do not, in general, go to the most qualified or talented scientists, hence the word "grantsmanship&
75.
by dnautics 1mo ago
if those scientists were to raise money from private donors would they find themselves in jail, or their property confiscated to where they couldnt continue research? that is the definition of "freedom of speech". if you accept m
76.
by dnautics 1mo ago
> outside of explicit training what does this mean anymore? i can invent a proof assistant language that was not in the training set and the llm does a fantastic job with it. (github.com/ityonemo/bpa) well it learned some sor
77.
by dnautics 1mo ago
since when does freedom of speech cover "getting money from the government". there is broad latitude for to government to restrict the speech of employees in general (if you are a nonappointed employee of the executive branch yo
78.
by dnautics 1mo ago
It's not just thought terminating. It's hypocritical why are you even doing in tech? your users should just use boring technology, pen and paper is great and boring, no need for a computer, much less a saas.
79.
by dnautics 1mo ago
i dont know. ive seen and dealt with so many shit codebases that were justified because python/java/etc are "boring technologies" that had extreme gunning fog index for understanding the code because those boring techno
80.
by dnautics 1mo ago
im not thinking. im an autocomplete with fat fingers (too lazy to fix my mobile keyboarf spelling misyakes)
81.
by dnautics 1mo ago
if you wanted to further study the phenomenon you could probably run the experiment again, and the ablate or corrupt those intermediate activations to get a feel for what it was thinking at the "time".
82.
by dnautics 1mo ago
see sibling comment, > see it actually improve just through accreting context this actually happens and has been tested.
83.
by dnautics 1mo ago
> to conclude if it is thinking or not I would want to look inside its head https://arxiv.org/abs/2607.03502 a non-thinking token model (just "completion") can answer one-step questions but generally not m
84.
by dnautics 1mo ago
nope. trade corrected us co is down considerably too. biggest contributor is that there was a flippening the us biggest trading partners shifted back to canada and mexico from china, under trump 1 iirc, which are way less bad emitters than
85.
by dnautics 1mo ago
no, we should actually put some teeth into this idea that this is a global problem and make everyone do their part. the us is dramatically reducing its co2 footprint. the us currently is emitting per capita roughly what it emitted oer capi
86.
by dnautics 1mo ago
> what evidence would convince you that it is thinking so, none it seems. as its behaviour becomes more and more humanlike you can just move the goalposts and say "thats consistent with an autocomplete" buddy i got some bad ne
87.
by dnautics 1mo ago
> t doesn't think in humans the exact same behaviour (cheating) is slmost always the result of a chain of complex series of choices and environment-driven rationalization. if the llm doesn't cheat, you say "its just produc
88.
by dnautics 1mo ago
just punish countries a fixed amount per kg yoy increase. small countries will (probably) not be punished much in case they had overage
89.
by dnautics 1mo ago
well theres a subset of countries that are reducing their emissions and a subset of countries that are not. perhaps we should punish the countries that are not.
90.
by dnautics 2mo ago
> I know this is obvious, but I never actually thought about it. I'm probably not the only one. It's pretty obvious, and I have thought about it, but I bet this is the sort of thing that I would reach for, forget about and lo

More ›