ithkuil

18 ms·

331.
by ithkuil 2y ago
well that really depends on the metric right? do you care about quality of life? literacy? about social justice? about women's rights? or about who has the most richest billionaires or the strongest military? or who has the best americ
332.
by ithkuil 2y ago
Yeah reasoning models are "self-doubt" models. The model is trained to encourage re-evaluating the soundness of tokens produced during the "thinking phase". The model state vector is kept in a state of open exploration.
333.
by ithkuil 2y ago
A good interview with him: https://www.samharris.org/podcasts/making-sense-episodes/346...
334.
by ithkuil 2y ago
I don't quite understand the logic. Even if you have cheaper models if you have tons of compute power you can do more things than if you had less compute power! You can experiment with huge societies of agents, each exploring multitude
335.
by ithkuil 2y ago
Perhaps because those things already happened at the beginning of the Chinese Revolution and we're just seeing a phase that's several generations on the future of the transition
336.
by ithkuil 2y ago
there are multiple success metrics
337.
by ithkuil 2y ago
Yeah that weird logic is all too common recently. I'll paraphrase: The ravenous wild beast is pounding at your door. The only thing standing before you and certain death is your front door that needs urgent repairs. You call your landl
338.
by ithkuil 2y ago
I think you may be the one in a bubble. I don't think the average conservative grandma is really in favour of the chaos that Elon is unleashing. The only reasons they voted for Trump is that they thought they were against total debauch
339.
by ithkuil 2y ago
It may be an acceptable compromise if the two alternative options weren't so extremely unpopular with the other side. That's why the checks and balances are required.
340.
by ithkuil 2y ago
Please read more carefully what I wrote. Most trump voters may well he on board but that doesn't mean that most of the overall voters are. A fraction of people who voted for Trump thought he would be like Trump 1.0 and just wanted to s
341.
by ithkuil 2y ago
Is it? A simple boolean choice and you give total carte blanche to anybody got elected? I think a good chunk of the voters cast their choice assuming that there would be checks and balances. Sure there are also many, many trump voters who r
342.
by ithkuil 2y ago
10 years ago I went on a similar journey. I left faang to work on a startup working on embedded firmware for esp8266. The lack of tooling was very frustrating. I ended up writing a gdb stub (before espressif released one) and a malloc debu
343.
by ithkuil 2y ago
The unison programming language ( https://www.unison-lang.org/ ) follows a similar idea. Functions have immutable IDs and "modifying'" a function involves creating a new function and all the callers that need t
344.
by ithkuil 2y ago
Damned if you do and damned if you don't
345.
by ithkuil 2y ago
The two worldviews can be reconciled: people are generally good but a relatively few bad people turn the world in a dangerous place that requires counterbalancing in order to bring the stability needed for the majority of good people to thr
346.
by ithkuil 2y ago
I think the idea is that black holes can never shrink to a size small enough to emit a hawking radiation hotter than CMB as long as it receives photons from the CMB that that will balance the mass/energy of the black hole and keep it i
347.
by ithkuil 2y ago
TIL it comes from: "1754: coined by Horace Walpole, suggested by The Three Princes of Serendip, the title of a fairy tale in which the heroes ‘were always making discoveries, by accidents and sagacity, of things they were not in quest
348.
by ithkuil 2y ago
A few minutes ago I picked up a dusty CD at random, after possibly multiple years of not listening to my CD collection. It was a REM album. Then I open HN and I see this title. Serendipity
349.
by ithkuil 2y ago
this is why the social media powered bullshit machine is so effective. Scenario: Chief executive president says that children are raped by unicorns. People react by saying, there is no way, unicorns don't exist. How do you know they do
350.
by ithkuil 2y ago
I think the idea is that fuzzy search makes it easier for users to spot typos and this is a general feature of the flag library not something specific for --help. Short-circuit any typo may be dangerous. Now, an argument can be made for inc
351.
by ithkuil 2y ago
It's probably fuzzy similarity search
352.
by ithkuil 2y ago
ZT came first
353.
by ithkuil 2y ago
I understand that anti-black racism is still a factor today. I really fail to see how doubling in on the importance and reality of race and skin color in particular (as opposed to culture) everywhere is going to make that problem go away.
354.
by ithkuil 2y ago
But that should be the goal. Using skin color as a proxy for being disadvantaged works only inasmuch the proxy is precise enough. If it's not, it's just another bias. If you had a better metric that can be used to help people fro
355.
by ithkuil 2y ago
We live on times of the great pendulum swings. It's hard to have nuanced conversations.
356.
by ithkuil 2y ago
Trump said so many things, so many outlandish things that I think a sizeable percentage of the people who voted for him didn't quite believe he would do. I don't have hard data on that but I did speak with several repubblican vote
357.
by ithkuil 2y ago
> You're hoping for a salacious answer. no I don't. I literally said it was a honest question. I truly just want to know more about the underlying mechanisms of why governments classify and declassify things because I don'
358.
by ithkuil 2y ago
But clickety click, Barba Trick: it's not formally a department, so there is now law stopping the president from creating this informal advisory thing. It's not clear exactly what it is. But they just put the word "department
359.
by ithkuil 2y ago
Honest question: why weren't they declassified earlier?
360.
by ithkuil 2y ago
I believe GP refers to the fact that they set out to measure the speed of the earth relative to the aether only to discover there was no such thing.

More ›