General > General Technical Chat
Is Arduino killing the electronic hobby?
langwadt:
--- Quote from: Zero999 on July 03, 2021, 09:42:48 am ---
--- Quote from: eti on July 02, 2021, 01:12:44 am ---Remember that song, "Video killed the radio star"? Maybe SHORT term, but look - where's MTV now? Nowhere, and where's music radio? EVERYWHERE. See my analogy?
--- End quote ---
I don't think either radio, or MTV are doing well, at the moment. They've both been replaced by streaming. I've never seen the point in music videos, or watched MTV in the 90s, when it was at its peak. Nowadays, when I listen to music, I put it on YouTube, in a background tab, so use the video part.
--- Quote from: Nominal Animal on July 02, 2021, 01:31:01 am ---Full disclosure: I just stumbled on a machine learning-colorized version of Nosferatu (1922), with a nice orchestra-based soundtrack.
--- End quote ---
Well AI has killed manual colourising to some degree. There's still some need for manual training, to colour thing such as clothes, old vehicles, buildings etc. but most of it can be automated.
--- End quote ---
https://www.reddit.com/gallery/mqmvx5
Zero999:
--- Quote from: langwadt on July 03, 2021, 10:03:22 am ---
--- Quote from: Zero999 on July 03, 2021, 09:42:48 am ---
--- Quote from: eti on July 02, 2021, 01:12:44 am ---Remember that song, "Video killed the radio star"? Maybe SHORT term, but look - where's MTV now? Nowhere, and where's music radio? EVERYWHERE. See my analogy?
--- End quote ---
I don't think either radio, or MTV are doing well, at the moment. They've both been replaced by streaming. I've never seen the point in music videos, or watched MTV in the 90s, when it was at its peak. Nowadays, when I listen to music, I put it on YouTube, in a background tab, so use the video part.
--- Quote from: Nominal Animal on July 02, 2021, 01:31:01 am ---Full disclosure: I just stumbled on a machine learning-colorized version of Nosferatu (1922), with a nice orchestra-based soundtrack.
--- End quote ---
Well AI has killed manual colourising to some degree. There's still some need for manual training, to colour thing such as clothes, old vehicles, buildings etc. but most of it can be automated.
--- End quote ---
https://www.reddit.com/gallery/mqmvx5
--- End quote ---
That pretty much agrees with what I said: manual training is required for things which the AI won't know the colours of such as clothing, but lots of it: sky, trees, grass, plain old asphalt and uncoloured buildings can be done automatically. If the author repeated the experiment, on a motion picture, after training the AI on those original stills, or samples of clothing, he/she would have an authentic result.
RoGeorge:
--- Quote from: fourfathom on July 01, 2021, 03:49:53 pm ---
--- Quote from: RoGeorge on July 01, 2021, 12:31:15 pm ---What we are looking for is, in fact, a human-like civilizations but from outer space.
--- End quote ---
What do you mean "we", kimosabe? Yes, perhaps the popular press and the "I Want to Believe" types are focusing on human-like aliens, but the serious researchers (as with the SETI Institute, where I used to be involved) are considering all types of life, including virtual and non-organic (as you mentioned). The only commonality this potential alien life necessarily has with humans is that it produce some signal or other evidence of existence that we can detect from the Earth locale.
--- End quote ---
By "we" I was thinking "us, humans" as a species. Individual mileage may vary.
The looking for "human-like aliens" put as "a kind of dumb thing to look for" was not about being stupid. I apologize if it sounded like that. My bad phrasing. I was just trying to highlight how small is the probability to detect aliens. In fact, while I was writing that I thought SETI ended many years ago, no way I would have aimed at SETI. (but I googled now and it seems SETI still exists, no idea where from I knew it was ended)
It was not about the pop culture green aliens and flying saucers either. The "human-like aliens" was about aliens close enough with us so we will be able to notice them (e.g. if a "meme", or the society, or the Internet were to be in fact aliens, we wouldn't know it even if we interact with them daily), and to have a meaningful interaction with them, meaningful from our standpoint (e.g. if it were to discover aliens like the thinking ocean from Stanislaw Lem's "Solaris", that would be more like a curiosity).
I believe we are strongly limited by the place where we evolved, limited not only by our biology, but by our mind, too. I don't think one day would be possible to know everything.
Our marvelous brain with which we took over over our animal kingdom, brain of which we are so proud of to have would be nothing if one day an alien with a brain the size of an ocean will one day show of. Size alone matters a lot. From size and big numbers alone can emerge things we don't comprehend. We still struggle to define what is conscience.
Another limitation come from the fact we can only see/detect/perceive only a very small part of what it is "out there", being it directly or with instruments. For example, only 10 generations or so ago we were unaware about radio waves. Now we are looking for radio signals assuming aliens would use them, too. Highly unlikely, but does this mean SETI should stop using radio-telescopes? No, of course we must look and search with what we have, no matter how limited we are now in tools or in understanding.
Another limitation comes from the training of our "General Intelligent" NN (by the way, AGI term is not mine, just that it seems nowadays AGI somehow became the de-facto name, meanwhile searched for AGI prize and it seems there are many, e.g. a $5 mil one). To give an example about the kind of limitation I picture, let's imagine a very simple NN trained for OCR (Optical Character Recognition), and we tech it to recognize only digits from 0 to 9. That's the only thing our OCR can, or knows, that's the only thing it can understand, that's its entire world. Now, if we suddenly show to our OCR a letter, the OCR will either misclassify the letter as a digit, or it might miss it entirely considering it's noise or an ink smudge. Same with us and very different aliens. That's why I wrote human-like aliens, or else we would misidentify them or wouldn't notice them.
We evolved and lived here on Earth for a very long time, and in very specific conditions. In all we do and think there is a gazillion of tacit assumptions. The very think that we try to reason using logic, and assuming true and false are opposite and can not overlap is already a limitation.
Sure, in math there could be fuzzy logic instead of boolean logic, but math is an axiomatic system and it's abstract. We can set as axioms whatever, but those axioms will have to come from our NN limited by size and training (unless we want to start from random axioms, which will be quite meaningless), so not even math can pull us completely outside of the box we are imprisoned now, the box of our own limitations.
For now we only know to listen for electromagnetic waves, and we learned how to do that only moments ago (at an astronomical scale). Maybe we know a couple of other few things than EM waves (thinking Fermi, CERN, LIGO) but not much else. This is almost like sitting in a medieval castle and hoping to see smoke signaling, but underwater. Could happen, but I wouldn't hold my breath waiting for it.
Then, there is always the small chance that we are the first and we are the Gods. We even created life already. (Thinking here the cloned Dolly sheep, or Craig Venter's team booting synthetic biologic life, or maybe the dude that created living assembly code lines long time ago - I don't recall his name, I think it was contacted later by SETI, too, or maybe it was NASA, not sure, his study was about how the amplitude of random changes between generations, and how that - the equivalent of what temperature is for us - can kill or give birth to small self-perpetuating assembly programs, thus proving life can only be expected to form only in a certain range of temperatures, too much agitation will dismantle it, too little will froze it dead - IIRC this was first tested with PDP assembly code)
What I like to think is that the Universe has much, much more "stuff" in it, and we will never be able to look at it other then through a keyhole. It's an impossibility to ever know it all.
Is this discouraging? No, it's entertaining because there are so many to find out about.
So what's the meaning of all this, then? What's the meaning of life, the meaning of existence itself? None. It's meaningless. We like to assign meaning, maybe so we can have a meaning, too, but there really isn't any. We insist in pretending there is a meaning, instead of growing up and admitting we are meaningless just like anything else.
Is this good, or bad? Well, there is no such thing as good or bad either. The notion of good and bad only makes sense in relation with a given goal. For example, it's good that I made that soup yesterday. Not for the chicken. It depends of the goal if something is good or bad, or in terms of NN it depends of the cost function we want to optimize for.
So, is Arduino good or bad? It depends of what's the goal when judging that.
fourfathom:
--- Quote from: RoGeorge on July 03, 2021, 06:37:04 pm ---I thought SETI ended many years ago, no way I would have aimed at SETI. (but I googled now and it seems SETI still exists, no idea where from I knew it was ended)
--- End quote ---
SETI research at NASA was defunded many years ago (this was in the Proxmire "Golden Fleece" era). In response, the SETI Institute was formed, funded by private individuals, Carl Sagan and Frank Drake among them. More recently, NASA has been working with the SETI Institute and other SETI groups on exobiology and extremophile research. There are many scientists who call the SETI Institute home, but get their funding from NASA as they participate in NASA projects. Some of the key scientists on the Kepler mission are SETI Institute folks. I was a Director at the Institute for a while, during the Allen Telescope Array project.
There are other SETI ventures. "SETI at Home" was developed at Univ California at Berkeley.
Yes, we use the tools we have. We (SETI) do hope that intelligent life, whatever form it takes, will want to communicate with life elsewhere, and that they choose one of the few modes that can propagate vast interstellar distances. Unless unknown physics are involved, the tools we have may be able to receive these signals. Of course, ETIs may not want to risk announcing their presence, because of the same concerns that many here on Earth hold. I think it highly unlikely that random "I Love Lucy" TV transmissions (etc.) can be received many light years distant, and in general as communications technology advances the signals look more and more like noise the window for accidental reception is vanishingly small. So it has to be a deliberate beacon, one designed to be received. Highly unlikely, but IMO worth the search if only for "human" reasons.
RoGeorge:
Thanks for taking the time to explain, and WOW! :-+
Please, I'm a sucker for firsthand history tales.
How was it there, at SETI?
Navigation
[0] Message Index
[#] Next page
[*] Previous page
Go to full version