Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
mach1ne
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
27 ms
·
301.
▲
by
mach1ne
4y ago
It shows how immature of a world we live in that ideas become automatically discredited at the highest levels depending on who's saying those ideas out loud.
302.
▲
by
mach1ne
4y ago
Didn't German Green Party greenlight (sorry) driving down nuclear energy in favor of coal? Political parties tend to minmax popular support over success of personal ideology, perhaps that's the case here as well.
303.
▲
by
mach1ne
4y ago
The people in my dreams are also often not proportional to the amount I see them in real life. I've found that I at least remember mainly dreams which had a strong emotional footprint, and teenage years definitely have the largest one.
304.
▲
by
mach1ne
4y ago
This is the most enlightening comment I've read about this case.
305.
▲
by
mach1ne
4y ago
But if there is no obviously desirable move, and instead a wide array of potential equally good moves, then picking the same as the computer picked gives a high probability of cheating. That being said, I find it unlikely that that's t
306.
▲
by
mach1ne
4y ago
Indeed, Stable Diffusion does not currently run on AMD graphics cards.
307.
▲
by
mach1ne
4y ago
For me it complains about NSFW when I ask a smirking expression on the face of the example prompt. I wonder if this is using some special NSFW filter?
308.
▲
by
mach1ne
4y ago
Humans may naturally be monogamous, at least to a degree. It is a common occurrence in the animal kingdom, in many bird species for example. Monogamosity is therefore a very unlikely candidate for fertility drop.
309.
▲
by
mach1ne
4y ago
A new weapon always begets new counter-weapons. They may provide sufficiently believable protection from the drones, though not all-encompassing. Wars may become somewhat bloodier and a lot scarier. In any case, I highly doubt drones would
310.
▲
by
mach1ne
4y ago
While impressive, it doesn't imply that GPT would have any significant 'task memory'. Remember that it always predicts the next token or word - as such, it essentially recognizes whether the next 'task' in the list
311.
▲
by
mach1ne
4y ago
OpenAI was the first on GPT-level text generation, as it was (I think) on DALL-E level image generation. It's winner takes all rather than marketing.
312.
▲
by
mach1ne
4y ago
The authors are seemingly falling into the fallacy of replicating the human experience, which simply leads to ultimately inefficient Alt Intelligence. While an interesting idea which could have a couple very rare useful use-cases, this seem
313.
▲
by
mach1ne
4y ago
To be precise, scaling improved results logarithmically. Accuracy seems to plateau with scaling and never reach any magical limit of true general intelligence.
314.
▲
by
mach1ne
4y ago
This actually harbors a good point. The quality stays high if the interactions users receive are strongly weighed towards the high-quality content/people.
315.
▲
by
mach1ne
4y ago
That's a good definition on consciousness. Not one non-materialists would share, but any entity which fills those requirements is likely indeed one which would deserve the materialist stamp of consciousness.
316.
▲
by
mach1ne
4y ago
Could you link some of these models? An interesting perspective that asymptote.
317.
▲
by
mach1ne
4y ago
First you have to define consciousness, and especially the external difference between a conscious and non-conscious intelligence.
318.
▲
by
mach1ne
4y ago
Indeed they demonstrated that, but don't underestimate the effect an existential crisis like this has. We will almost certainly see heavy changes coming to the Russian military. It could be that in five to ten years their military is a
319.
▲
by
mach1ne
4y ago
I should add that Russia desires the Baltics, which are in an awkward position to defend by NATO. Finland has a strong military and its proximity to those countries makes an incursion somewhat less likely.
320.
▲
by
mach1ne
4y ago
It doesn't, not really. It ultimately just creates an illusion of agreement due to faster transmission.
321.
▲
by
mach1ne
4y ago
It's interesting how in the olden times large cities were almost like an entity of their own. Their stories significantly differed from other cities and their reactions to odd phenomena such as one in the article were unique and diffic
322.
▲
by
mach1ne
4y ago
>but it's not expensive _everywhere_. It's expensive in some of the most desirable places. The fact that the median and average house prices have exploded signifies that the housing market is in a bubble. It's not relevant
323.
▲
by
mach1ne
4y ago
However, the study concluded that person's ability to produce randomness peaks at 25. An increase in showing bad faith doesn't tell us anything about the ability to produce randomness if desired. Thus, if we accept the bad faith a
324.
▲
by
mach1ne
5y ago
I don't know if split-brain studies are familiar. In short, some people had the connections between their lobes cut off to combat epilepsy. As a result they had two separate brains, and if, for example, a funny image was shown to their
325.
▲
by
mach1ne
5y ago
Can you emulate personalities? Can you, for example, ask Elon Musk for an opinion on something you're currently working on? If you can, a routine discussion with internal monologue is most likely more of a habit than a neurological fea
326.
▲
by
mach1ne
5y ago
Interesting! Gauging from the original Reddit post, it seems that people exist on a spectrum regarding this for whatever reason. This could explain a lot about why some people learn certain stuff faster than others. For example, maths encod
327.
▲
by
mach1ne
5y ago
I also deduced that my inner monologue isn't making any sound even though I 'hear' the words. For each word it's like the average of every time you have heard that word. But it's easy to mistake that into hearing th
328.
▲
by
mach1ne
5y ago
> rather than waiting for a central authority, financed with other people's money, to give up on one and pick up the next when it finally gets over its ego. Arguably, currently the ego problem is very real with the individuals who h
329.
▲
by
mach1ne
5y ago
> if it's correct, smart people will eventually be convinced of that. That's not how science has historically worked. Smart people won't be convinced, new people will. Status quo changes when the previous generation is lai
330.
▲
by
mach1ne
5y ago
While neurological problems are an undeniable and perhaps the scariest aspect of COVID infection (aside from death), I wonder if sometimes the problems with word recall are simply caused by the long-term lack of social contacts. My own soci
More ›