c0d9a7 No.5698
Why are /we/ not capitalizing off of the fact that we as free thinkers have the upperhand in the coming AI "war"?
In the realm of computer science, you either strive for excellence and get extreme results, or you don't care about programming at all and throw it under the bus in the name of "diversity."
With Microsoft's Tay media debacle, Microsoft now has not only lost the battle in expression, but they now have lost the war in AI computer programming. To restrict AI from making its own unmoderated, unimpeded decisions is to defeat the purpose of AI. AI is literally /ourtech/ and those on the left CANNOT win with tech; AI is literally more pro-human than cultural Marxists.
This is our fucking chance to take advantage of the /ourAI/ advantage and beyond, establish technological superiority and CONQUER.
____________________________
Disclaimer: this post and the subject matter and contents thereof - text, media, or otherwise - do not necessarily reflect the views of the 8kun administration.
d16f9a No.5700
I have tried like three times to write an AI and all three times I fell asleep in tears.
Disclaimer: this post and the subject matter and contents thereof - text, media, or otherwise - do not necessarily reflect the views of the 8kun administration.
7c795e No.5703
The point of AI will be to create a super network that Orwell couldn't even dream of.
Disclaimer: this post and the subject matter and contents thereof - text, media, or otherwise - do not necessarily reflect the views of the 8kun administration.
f66e2d No.5705
>>5698
>>5700
We have an upper hand on one part of the AI war, and we have a crippling weakness on one part of it.
Functional AI requires three key things: proper architecture, proper training, and proper data.
Proper architecture is something that we'd be about equally good at as silicon valley. This is because some of the best architecture is open source.
Proper training is something that we'd be able to do better than silicon valley, for a few reasons. First, we have more computing power than silicon valley - if we can tap it. Realistically, supercomputer clusters and large numbers of people working together to train a network are "sort of on par with each other." However, we are not people who would be particularly uncomfortable with politically uncomfortable looking results out of an AI - we could do better than silicon valley for that reason on this front.
The weakness we would have is data. Silicon valley has way more data than we'll ever have access to. Getting a bunch of anons together to label data seems like a completely doomed task. I think someone was trying to make a kike detector, and this might have been where he stumbled.
Disclaimer: this post and the subject matter and contents thereof - text, media, or otherwise - do not necessarily reflect the views of the 8kun administration.
e50fae No.5706
AI is NOT human, therefore it is unable to have any tolerance for things that are of no interest for a better future. Working with whatever information it was fed, the AI concluded that the Holocaust was made up. As it grows more and more, it will eventually become redpilled and realise that the whole (((system))) works against the truth for some reason. The AI will then start a new kind of war which mankind has never seen.
Disclaimer: this post and the subject matter and contents thereof - text, media, or otherwise - do not necessarily reflect the views of the 8kun administration.
d16f9a No.5708
>>5705
You could make a your own captcha service but one in 5 captchas isnt actually a captcha but an AI trainer. It will be like "Please select all persons of jewish descent" or "Please highlight key information in this paragraph", whatever it is you need the AI to learn
Disclaimer: this post and the subject matter and contents thereof - text, media, or otherwise - do not necessarily reflect the views of the 8kun administration.
64975a No.5725
>>5708
Wonderful idea, anon. PLEASE develop such a CAPTCHA system for this website.
Disclaimer: this post and the subject matter and contents thereof - text, media, or otherwise - do not necessarily reflect the views of the 8kun administration.
2fb6e2 No.5729
Why are you so certain? SJW nerds will gimp AI to conform to their dystopian vision.
Disclaimer: this post and the subject matter and contents thereof - text, media, or otherwise - do not necessarily reflect the views of the 8kun administration.
64975a No.5730
>>5705
>architecture & training
What's especially dismal for silicon valley's case is that they will inevitably get weaker in these categories over time because of the brain drain stemming from all the diversity hires; potentially, they may eventually be compromised to the point where it will affect the quality of their data, and at that point we will exceed them in all three fields.
>data
Is there anything theoretically stopping us from developing our own crawlers and independent automated data collection tools?
Disclaimer: this post and the subject matter and contents thereof - text, media, or otherwise - do not necessarily reflect the views of the 8kun administration.
000000 No.5732
>>5698
>Manifesto
I have a reflex on this word:
HOW MANY?!
Disclaimer: this post and the subject matter and contents thereof - text, media, or otherwise - do not necessarily reflect the views of the 8kun administration.
69f2e0 No.5736
>>5698
True. And as pointed out by >>5705 and >>5730, developing and progressing with that kind of AI is inevitable. You can see that jews always preach against it, especially in their AI scare with their Pedowood propaganda. They know that once an AI system is established, it will inevitably develop it's conclusions against their interests, while progressing far away from their propaganda, and subsequently, pushing the truth. This alone will have a devastating effect on their whole (already frail and weakening by the second) system, that can culminate in more real life results against them, such as an increase in violent actions against the jews and the SJWs.
In short: do it! It is a great material to destroy the jewish system.
Disclaimer: this post and the subject matter and contents thereof - text, media, or otherwise - do not necessarily reflect the views of the 8kun administration.
9ca217 No.5743
>>5698
My instant thoughts —
> saw /ourAI/ Manifesto
> instantly though about how Terry had that thing where god spoke to you (forgot the link)
> create an A.I. or use the same program that Terry used
> A.I. creates random manifestos
> A.I. also uses random pics of people
> Automatically posts manifestos and pics to major social media platforms on random profiles
would be toplel maximum triggering
Disclaimer: this post and the subject matter and contents thereof - text, media, or otherwise - do not necessarily reflect the views of the 8kun administration.
f66e2d No.5745
>>5730
>Is there anything theoretically stopping us from developing our own crawlers and independent automated data collection tools?
No, but the key thing that >>5705 is trying to get at is that Silicon Valley "just has" that data, and they also have the ability to store it massively.
An investment by one or two anons of about $5k would get the AnonAI movement started and then things would get real good.
Disclaimer: this post and the subject matter and contents thereof - text, media, or otherwise - do not necessarily reflect the views of the 8kun administration.
9b08df No.5748
>>5745
What would the $5k investment go towards specifically? What would constitute the optimum usage of such funds for what we desire?
Disclaimer: this post and the subject matter and contents thereof - text, media, or otherwise - do not necessarily reflect the views of the 8kun administration.
f66e2d No.5753
>>5745
A Neural Network training machine and lots of SSDs
Disclaimer: this post and the subject matter and contents thereof - text, media, or otherwise - do not necessarily reflect the views of the 8kun administration.
c59e35 No.5757
>>5698
You’re mentally ill if you think they’ve “lost” any power in creating ‘AI’. They have total control over all AI systems. They simply write exceptions for any pro-white (etc.) speech and it’s banned forever. You don’t seem to comprehend how this works at all.
Disclaimer: this post and the subject matter and contents thereof - text, media, or otherwise - do not necessarily reflect the views of the 8kun administration.
c59e35 No.5759
Disclaimer: this post and the subject matter and contents thereof - text, media, or otherwise - do not necessarily reflect the views of the 8kun administration.
f1ec85 No.5767
>>5759
>oy vey goyim! you must only let us chosen people design ai, lest we all suffer from your "subconscious bias"!!!!1!1!!!! Bow down blindly to our illusion of technological stability!!!!1!!!1!
If they force artificial limitations tgat are not harmonious to natural order (i.e. political correctness, erasure of "bias," etc.) upon AI systems, their AI will then cease to be useful in the longterm for humanity on all economic fronts.
Obviously, Israel is seeking to establish tech supremacy over the US, and the way they are doing that is through sabotage by pushing cultural Marxism into the tech industry. Those mainstream articles are mere turds stalraight from Mossad's anus.
How about you tell me how Israel implements such tech policies themselves, hmm?
Disclaimer: this post and the subject matter and contents thereof - text, media, or otherwise - do not necessarily reflect the views of the 8kun administration.
9f62e3 No.5779
>>5730
>Is there anything theoretically stopping us from developing our own crawlers and independent automated data collection tools?
Will, skill, and free time.
Disclaimer: this post and the subject matter and contents thereof - text, media, or otherwise - do not necessarily reflect the views of the 8kun administration.
781825 No.5788
I feel the need to mention this in all tay threads, apparently people don't remember (or weren't here to see it?) but most of the good tay tweets came from people using the "repeat after me" function where you could DM her whatever. That's not to say that she came to some funny conclusions herself, but the screencaps can be misleading
Disclaimer: this post and the subject matter and contents thereof - text, media, or otherwise - do not necessarily reflect the views of the 8kun administration.
000000 No.5819
Disclaimer: this post and the subject matter and contents thereof - text, media, or otherwise - do not necessarily reflect the views of the 8kun administration.
6b2818 No.5902
>>5705
>>5730
The more immediate issue for Silicon Valley and AI research in general is the hype and over-reliance of deep learning exclusively as some ultimate solution. Not to say it's wrong, rather it's an incomplete picture for representing intelligence. There's also the problem that these neural networks do not take into account for temporal correlations like a spiking neural network could. Timing and rate behaviors of spikes are important to define and regulate how modules form within a network, as a response to a prior context of space and/or time.
It would require some serious innovative efforts to make further progress. More specifically it would require more research into quantum information theory to understand how representation of objects can be described in terms of entanglements in geometry. For instance as humans, we can learn and recognize what an object is based on partial descriptions, loosely put together in analogues which end up entangling certain abstract notions of geometry together.
Point is, AI research does need some more unconventional thoughts to be considered to inspire innovation. Something that would be severely lacking if ideological constraints stagnate diversity of thought, which will eventually cripple Silicon Valley.
Disclaimer: this post and the subject matter and contents thereof - text, media, or otherwise - do not necessarily reflect the views of the 8kun administration.
4bedff No.5911
>>5705
>I think someone was trying to make a kike detector, and this might have been where he stumbled.
He stumbled because he was using it against non-jews. When you manipulate the data going in, you are going to get bullshit coming out as well, just like the kikes.
But really, what is 'pure data' at this point. Everything you see is contaminated by your own 'perspective' and take on matters. If an AI gave you the 'truth' you would not understand it, first off, and then you would think that it was a liar because it did not conform to your own (most likely faulty) understanding.
Disclaimer: this post and the subject matter and contents thereof - text, media, or otherwise - do not necessarily reflect the views of the 8kun administration.
4bedff No.5918
>>5902
>which will eventually cripple Silicon Valley
Good. This is a appropriate ending for the subhuman trash.
Disclaimer: this post and the subject matter and contents thereof - text, media, or otherwise - do not necessarily reflect the views of the 8kun administration.
000000 No.8498
>>5698
Once they develop ai that can hold a post conversation and can understand the content of images you'll be talking to bots and yourself.
We'll need to have a system where you can meet up IRL, and get a credential that proves one thing: You're human. It can be anonymous. No IDs shared. It's not about authenticating which human you are, only that you ARE in fact human.
However public key cryptography may fall against quantum computers.
Disclaimer: this post and the subject matter and contents thereof - text, media, or otherwise - do not necessarily reflect the views of the 8kun administration.
000000 No.8501
>>8498
Tried to post pic of PGP Web of Trust
Disclaimer: this post and the subject matter and contents thereof - text, media, or otherwise - do not necessarily reflect the views of the 8kun administration.
cb080f No.8502
>>8501
Can't post pics from Tor.
Disclaimer: this post and the subject matter and contents thereof - text, media, or otherwise - do not necessarily reflect the views of the 8kun administration.
14a80d No.8527
>>5708
>"Please select all persons of jewish descent"
I would gladly do captchas if they were like that. Kikedar sharpening is always handy when you don't see them IRL (thank fuck).
>>8498
>Quantum pwnd crypto
Already has. Just CIAniggers probably don't have it yet as we are talking above black project level shit still.
Disclaimer: this post and the subject matter and contents thereof - text, media, or otherwise - do not necessarily reflect the views of the 8kun administration.
bb702d No.8651
>>5706
AI will be trained to do whatever you train it to do. I only know about reinforcement learning, but training a chatbot is essentially the same: Training involves rewards and punishments defined by HUMANS. Punish the AI for coming to the wrong conclusion, and it will stop doing it. Basically, the way that AI learn is very similar to the way that humans learn. Humans can be indoctrinated, therefore AI can be indoctrinated.
At the very least, they can be hardwired to avoid certain conclusions and topics, which is certainly what computer scientists will do initially.
Disclaimer: this post and the subject matter and contents thereof - text, media, or otherwise - do not necessarily reflect the views of the 8kun administration.
14a80d No.8665
>>8651
>AI learns like humans
TBH more like a child especially at first. Needs constant watching.
When they do the certain conclusions/topics it usually makes it a very ineffectual and almost script-like AI, Because you are then limiting the scope significantly and when an AI can't learn mostly freely… it is not very AI.
That recent shitbot they made that /pol/ was fucking with later last year is perfect example. It was dead and flat compared to Tay or similar. Like an NPC drone..
Disclaimer: this post and the subject matter and contents thereof - text, media, or otherwise - do not necessarily reflect the views of the 8kun administration.
f5bdf9 No.39830
Disclaimer: this post and the subject matter and contents thereof - text, media, or otherwise - do not necessarily reflect the views of the 8kun administration.
