|
Cingulate posted:Neural nets/deep learning happens on the CPU. Everyone is looking at NVIDIA, and so far they're delivering. Inference happens on CPUs, but the learning is almost all GPUs these days. There are various other specialization approaches as well.
|
# ¿ Dec 3, 2016 04:37 |
|
|
# ¿ May 14, 2024 18:07 |
|
Do you not think that, f.e., LSTMs or memnets represent meaningful advances?
|
# ¿ Dec 22, 2016 23:14 |
|
Cingulate posted:And Memnet is in a totally different category from LSTMs. Yes, I know. They were two examples.
|
# ¿ Dec 23, 2016 01:04 |
|
I was asking a question in earnest -- did he not consider them to be meaningful advances. I wasn't asserting anything about what he should believe. Sometimes a question is just a question, not a Socratic feint.
|
# ¿ Dec 23, 2016 02:41 |
|
They're two things that I think of as being modern advances, that didn't seem to fit his description. I apologize for imperfect wording of my question. Perhaps I shouldn't post quickly from my phone in threads discussing such sensitive matters. Your accusation of bad faith posting seems disproportionate, and in bad faith itself. Memnets turn up more in the literature as "memory networks" it seems; Weston was primary on the original paper. I always heard them called memnets where I worked.
|
# ¿ Dec 23, 2016 03:01 |
|
Elias_Maluco posted:A whole lot of movies, cartoons, sci-fi books and comic books did pictured flying cars during the 80s and 90s Picturing them as signs of a fantastic future isn't really the same as being part of a thoughtful forecast of upcoming technology.
|
# ¿ Dec 24, 2016 12:59 |
|
Yes, the goalposts are on wheels.
|
# ¿ Jan 10, 2017 16:08 |
|
Thalantos posted:Like it strikes me that since it is about fooling a human into thinking it's a human, that already happens. I don't think that's actually what the test was.
|
# ¿ Jan 10, 2017 20:21 |
|
Thalantos posted:What was it? I guess you can frame it that way. It wasn't intended to be adversarial, by my recollection.
|
# ¿ Jan 10, 2017 21:25 |
|
Thalantos posted:How would it be adversarial? Is it because your comparing the AIs attempt to pretend to be human versus the human ability to tell it is "lying"? In that the judge isn't involved in the conversation, and the other participant isn't trying to expose the computer (or disguise themselves as a computer). Not that the other participant should know if it's dealing with a machine, of course.
|
# ¿ Jan 10, 2017 21:43 |
|
I think the Turing test is usually framed as being written, because perfect voice synthesis is hard and not really relevant to the core question. Would you consider a machine to have passed if it couldn't be distinguished from a 5-year-old? Someone with dementia?
|
# ¿ Jan 10, 2017 22:26 |
|
Cingulate posted:Humans don't learn the knowledge required to uphold civilization without supervision, so your definition is too strong. I've heard ravens are lovely with climate control.
|
# ¿ Jan 10, 2017 23:47 |
|
thechosenone posted:Kindof? Like if I it really felt like I was talking with a 5 year old I would be pretty impressed at least. Not nearly as much as if it were an adult, but still impressed. How about a 2-year-old? A written, detailed log of a newborn's activities? Or, I think more interestingly, a person with dementia or other serious neurological damage? I wouldn't say that such people have ceased to be intelligent, personally. That's not a philosophical line I'm comfortable crossing.
|
# ¿ Jan 11, 2017 01:20 |
|
I'm trying to figure out how much people want to test "is intelligent" versus "behaves like a human", and in the latter case what it means for a given capability to not be present for a human.
|
# ¿ Jan 11, 2017 02:41 |
|
|
# ¿ May 14, 2024 18:07 |
|
A Wizard of Goatse posted:people don't generally figure that things that don't look like people or communicate like people are people Yes, quite. Are we talking about artificial intelligence, or artificial people?
|
# ¿ Jan 11, 2017 03:44 |