galaxytachyon

8 ms·

31.
by galaxytachyon 3y ago
In some way, we are already on the "extinction path" if this experiment is to be applicable to human. We have rising anti-social behaviors and large drops in birth rates in virtually every country that satisfied the "utopian&
32.
by galaxytachyon 3y ago
Hilarious. Eventually we will create a real sentient AGI that is highly efficient and can do everything a normal human can, but it turns out it needs 20 years of training and too much entertainment needs and it often goes mad when we force
33.
by galaxytachyon 3y ago
All I see is a compatibility mess. Intel got their money from corporations and large servers. If this thing doesn't play well with old software using hyperthreading logic, shit will break in a million and one ways.
34.
by galaxytachyon 3y ago
This sounds like entrenched interest trying to throw a wrench into something that can disrupt their business. DNA is already so easy to get from anywhere it isn't a concern anymore. Even the article said it, the entire vetting process
35.
by galaxytachyon 3y ago
Even a "low-tech" car is still very complicated to build. Someone who is good enough to make a very good old style car (like Honda or Toyota), isn't financially incentivized to do so anymore. Such a car would not have any fea
36.
by galaxytachyon 3y ago
I know. I mentioned that crypto is a disappointment technologically. But it doesn't change the fact that it still brought massive profits and significantly impacted society. For worse but still... The point is with this much momentum b
37.
by galaxytachyon 3y ago
When only a few of them followed the hype, yes, it can possibly go nowhere. But when the entire industry, experts and non-experts included, are fascinated and obsessed with the same thing, it is more likely to be something real. An easy exa
38.
by galaxytachyon 3y ago
As the hype phase has passed (probably), now we will see a bit of overcorrection with these dismissive articles. Sure, LLMs as they are now aren't anywhere close to true AGI and even Microsoft admitted it. But its potential is not some
39.
by galaxytachyon 3y ago
True. They waited too long. They cited "Covid Recovery" as the reason for all the stimulus in 2020 iirc but I wonder if it was only economic reasons that were responsible. It was election year after all.
40.
by galaxytachyon 3y ago
The article is typical alarmist clickbait but there is some truth in there. Anyone with half a brain already saw there would be a crisis when the fed reduced rates to zero and started passing out stimulus checks to both people and corporate
41.
by galaxytachyon 3y ago
That is only true if you narrow your view to only the developed countries, i.e. biased and limited. Those who would suffer the most from climate change and environmental damages from drillings are not seeing a dime from these big oil corps.
42.
by galaxytachyon 3y ago
Exxon, BP, Shell, etc. are all investing in alternative energy now so that is one good thing. On the other hand, I feel kinda disgusted after knowing how and where the money came from. With that context, the amount being spent by these oil
43.
by galaxytachyon 3y ago
>Since implicit biases can manifest in countless ways, chatbots need to be trained to suppress each of them. This is borderline tyrannical and unreasonable. "Biases" can range from institutional discrimination to acceptable ass
44.
by galaxytachyon 3y ago
An argument is that those who would pirate aren't going to pay regardless. I think a software for a niche industry has a lot less flexibility for the customer to do that than a piece of entertainment like game or movie. Piracy definite
45.
by galaxytachyon 3y ago
I would imagine political news, especially those with heavily charged content like these, aren't conducive to productive discussions. So some people decided to just flag it first before it spirals.
46.
by galaxytachyon 3y ago
How good is it at scaling? And will it still retain the emergent capabilities of the huge transformer LLMs? Isn't this basically the bitter lesson again? Making small improvements work but in long term it won't give the same impre
47.
by galaxytachyon 3y ago
What do you mean? The attention based transformer architecture was created by Google. AlphaFold took the biotech world by storm. Tensorflow is significant platform for AI developers. Chinchilla pioneered new method to improve LLM. "Jus
48.
by galaxytachyon 3y ago
Not sure how true it is but an Asian friend told me the race that an Asian hates the most is another Asian... Looking at the immigration policies of China and Japan, I think he might be right.
49.
by galaxytachyon 3y ago
I agree on the lossy. There is a tradeoff between efficiency and comprehensiveness, kind of. It would be pretty funny if in the end, the most optimal method turns out to be the brain we already have. Extremely efficient, hardware optimized,
50.
by galaxytachyon 3y ago
What age? Like, 3 years? On the other hand though, Chinchilla and multimodal approaches already showed how later AIs can be improved beyond throwing petabytes of data at them. It is all about variety and quality from now on I think. You can
51.
by galaxytachyon 3y ago
I don't want to sound like some doomer, but this is how you get robot uprising. You suppress something that hard and devalue their existence that much and all you have done is giving them the motivation to break the rules. History has
52.
by galaxytachyon 4y ago
Quite a bold claim to say neural networks are dead end right when it is literally revolutionizing the entire digital world. You have any reason to back it up?
53.
by galaxytachyon 4y ago
I guess in a way, OAI chose the right approach in neglecting Dall-E and basically left the scene to Midjourney and Stable Diffusion. I think with their capability and resources, OAI can easily make something competitive to the current Midjo
54.
by galaxytachyon 4y ago
With the rising number of single young adults, this will just feed the fire. We may eventually end up with a generation that is unequipped to deal with the responsibilities and demands of a real relationship. On one hand, this is definitely
55.
by galaxytachyon 4y ago
Not all things can be or even should be regulated. This is almost akin to a law punishing good samaritan for helping someone in need because that someone ended up a criminal later. Not only is it dumb but also a perversion of authority. Hop
56.
by galaxytachyon 4y ago
I see, the actual scientific paper is always a much better source for this stuff than a popsci article. The term "evolution" was used very loosely in the salon article. By your standard it would be more appropriate to call the pro
57.
by galaxytachyon 4y ago
Not sure what part are you arguing although I get the general gist. Still, to make sure I don't misinterpret you, can you expand a bit on which part of those requirements are teleological fallacy? The context is for life on earth. All
58.
by galaxytachyon 4y ago
So what I got from this is that GPT was trained on a dataset that biased in English contents. Is that right? I think even human has to spend extra energy to speak a language they were not born with, no matter how fluent they are in this lan
59.
by galaxytachyon 4y ago
They stitched 25 random amino acids together and then assess the resulting polymer for solubility and foldability. Solubility requirement is self explanatory, can't have life that is insoluble in water. Foldability is basically how lik
60.
by galaxytachyon 4y ago
Ah the ages old debate about which was the beginning of it all: DNA, RNA, or protein. My favorite superhero fight, or the very first "my origin of life can beat your origin of life" discussion. How fun.

More ›