Re: [agi] Re: Finding a Simple Distribution In a Binary String

theres so many 1's and 0's there it must be a quantum state!!! -- Artificial General Intelligence List: AGI Permalink: https://agi.topicbox.com/groups/agi/T07349206c4d4db02-Mb26302b11b346ce16196afcd Delivery options:

Re: [agi] Re: Finding a Simple Distribution In a Binary String

[agi] Re: Finding a Simple Distribution In a Binary String

A Short Introduction to Entropy, Cross-Entropy and KL-Divergence: https://www.youtube.com/watch?v=ErfnhcEV1O8=6s -- Artificial General Intelligence List: AGI Permalink: https://agi.topicbox.com/groups/agi/T07349206c4d4db02-M00cbc0175f8c5584691eedb9

[agi] Whats everyones goal here?

I want to make sport robots,  so I can make something I wasnt good at pointless for everyone else to do. Whats u guys hitchup? -- Artificial General Intelligence List: AGI Permalink:

Re: [agi] International AI Issues

That movie looks like so much fun!!!  If I saw it when I was little I would have been in a strange fantasy world!!! -- Artificial General Intelligence List: AGI Permalink: https://agi.topicbox.com/groups/agi/T48b3b969c849a2e5-M1887f06c4839909b00ef25b0

Re: [agi] International AI Issues

Don't forget the gun is good: https://www.youtube.com/watch?v=TVakHZp5ZBE -- Artificial General Intelligence List: AGI Permalink: https://agi.topicbox.com/groups/agi/T48b3b969c849a2e5-M5ec2383e22ef772583c74ec6 Delivery options:

Re: [agi] International AI Issues

Need an Airsoft version... -- Artificial General Intelligence List: AGI Permalink: https://agi.topicbox.com/groups/agi/T48b3b969c849a2e5-M278f7dfc9bf02b0433b19a01 Delivery options: https://agi.topicbox.com/groups/agi/subscription

Re: [agi] International AI Issues

Yes,  that is indeed the problem when half your brain is on the ground behind you. But id like to see a boxing robot end up just as dangerous as one with a gun,    then even not even arming them they still present a problem to the young developing scientist.

Re: [agi] International AI Issues

I ran into this interesting video of what another AI developer has made. Everyone in this business needs to think about being on the receiving end of THIS technology. https://youtu.be/RxBa5bQfTGc BTW, for $600 you can buy one. Steve On Thu, Oct 10, 2019, 6:20 AM John Rose wrote: > On

[agi] Re: Finding a Simple Distribution In a Binary String

Thanks Keghn for that, very interesting. -- Artificial General Intelligence List: AGI Permalink: https://agi.topicbox.com/groups/agi/T07349206c4d4db02-M0a1f321c7cb9188df32f303c Delivery options: https://agi.topicbox.com/groups/agi/subscription

[agi] Re: Finding a Simple Distribution In a Binary String

keghnfeem 10:55 AM (4 hours ago) Actions Huffman coding\n for data compression.     nibble system https://www.techiedelight.com/huffman-coding/   |        /\ 

Re: [agi] Re: Noise and Confirmation Bias

exactly. -- Artificial General Intelligence List: AGI Permalink: https://agi.topicbox.com/groups/agi/T8c8ee84b385720a5-Mdd4a6fd68490be0179c46ba6 Delivery options: https://agi.topicbox.com/groups/agi/subscription

Re: [agi] Re: Noise and Confirmation Bias

 They have has some luck. Did Google Just Achieve 'Quantum Supremacy'? https://www.youtube.com/watch?v=kyuf1zt-TpU  But for computing they are doing all wrong. If the want to do it right they should  do in base on my superlumanit processor. Q bits is not a good way of storing information.  My

Re: [agi] Re: Speedrun continues

Yeah, not bad On Wed, 9 Oct 2019 at 23:23, wrote: > Name it something that stands for its ease of use. EasyPython > *Artificial General Intelligence List * > / AGI / see discussions + > participants

Re: [agi] Re: Noise and Confirmation Bias

soorrryy... i apologize...  i mean "stupid" women,  not women.  sorry lana and soph. -- Artificial General Intelligence List: AGI Permalink: https://agi.topicbox.com/groups/agi/T8c8ee84b385720a5-M08096e1e4ccee2f1b564393f Delivery options:

Re: [agi] Made a corner detector

Interesting - thanks. On Thu, 10 Oct 2019 at 22:56, wrote: > 3 shaders sorry. > *Artificial General Intelligence List * > / AGI / see discussions + > participants + delivery >

Re: [agi] Re: Noise and Confirmation Bias

Keghn,  P vs NP is on the cusp of the possible,  it may or may not be,  but its a place you can go as a programmer if its true,    note all quantum computers on the internet are fake...  they are just retarded theories probably started by women. --

Re: [agi] Made a corner detector

3 shaders sorry. -- Artificial General Intelligence List: AGI Permalink: https://agi.topicbox.com/groups/agi/Tc5e7e8b165534b2f-M0190f62e311f4330d8ab7b92 Delivery options: https://agi.topicbox.com/groups/agi/subscription

Re: [agi] Made a corner detector

heres 2 shaders, i do it in direct compute hlsl, so it gets to be real lazy code, cause its launched onto 2000 cores, so the imperitaveness to optimize is a lot more lax. the top filter is what i use to suppress,  the bottom filter is a binary clamped version. and the shader below that is

Re: [agi] Re: Noise and Confirmation Bias

Brownian nose. Ya, try to compress that. Distribution can be done with a traveling window nested traveling window algorithm. One for your seed size, bit, nibble, byte, word, or other. The other is size of the window or the amount of date taken in at a time. This will give a model of many levels

Re: [agi] Re: Noise and Confirmation Bias

apples can be bananas but bananas can only be apples if apples are bananas. -- Artificial General Intelligence List: AGI Permalink: https://agi.topicbox.com/groups/agi/T8c8ee84b385720a5-Mc6887ed69bbd2cdd8241f331 Delivery options:

Re: [agi] Re: Noise and Confirmation Bias

well lossy can be lossless but lossless can only be lossy when lossy is lossless... -- Artificial General Intelligence List: AGI Permalink: https://agi.topicbox.com/groups/agi/T8c8ee84b385720a5-Mb0eaf38d8a3424c8f3fa4748 Delivery options:

Re: [agi] Re: Noise and Confirmation Bias

John Rose, thanks for the wonderful poetry gleaming in my eyes -- Artificial General Intelligence List: AGI Permalink: https://agi.topicbox.com/groups/agi/T8c8ee84b385720a5-M22918dcbe874872f8df6fa37 Delivery options:

Re: [agi] Re: Noise and Confirmation Bias

yeah it doesn't matter they both have their place and can be mixed though there is some crossover but that might just be special cases of each... still makes you wonder though :)  Why invite distraction -- Artificial General Intelligence List: AGI

Re: [agi] Re: Speedrun continues

No idea what this is all about, but I'll take the compliment On Thu, 10 Oct 2019 at 21:54, wrote: > did u already know that one? otherwise u r smarter than I THOUGHT > *Artificial General Intelligence List * > / AGI / see discussions

Re: [agi] Re: Speedrun continues

did u already know that one?  otherwise u r smarter than I THOUGHT -- Artificial General Intelligence List: AGI Permalink: https://agi.topicbox.com/groups/agi/T0abfcf4b6bc70e98-Ma51ff7ab35a20d331fe2a616 Delivery options:

Re: [agi] Re: Speedrun continues

I get Pybol and Cobon... although - COBOL? Or what is the reference? lishing = eng-lish- ...? On Thu, 10 Oct 2019 at 21:46, wrote: > lishing > *Artificial General Intelligence List * > / AGI / see discussions + >

Re: [agi] Re: Speedrun continues

lishing -- Artificial General Intelligence List: AGI Permalink: https://agi.topicbox.com/groups/agi/T0abfcf4b6bc70e98-Maccd48629b5b31bacf16ff37 Delivery options: https://agi.topicbox.com/groups/agi/subscription

[agi] Re: Finding a Simple Distribution In a Binary String

1 word becomes 2 words. -- Artificial General Intelligence List: AGI Permalink: https://agi.topicbox.com/groups/agi/T07349206c4d4db02-M562058aa7ad5efaf1fa948bd Delivery options: https://agi.topicbox.com/groups/agi/subscription

[agi] Re: Finding a Simple Distribution In a Binary String

yes, one takes for granted that we separate into tokens because of the spaces between? So you need to make these distinctions there,  which is counting logic... where one adds the space, a 1 turns to a 2. -- Artificial General Intelligence List: AGI

Re: [agi] Re: Noise and Confirmation Bias

lossless and lossy are logically opposite, let alone any vernaculaic use. -- Artificial General Intelligence List: AGI Permalink: https://agi.topicbox.com/groups/agi/T8c8ee84b385720a5-M2cb43c3465401681f903556f Delivery options:

Re: [agi] Re: Noise and Confirmation Bias

Well at the time I thought I had theoretically developed a general lossless compression algorithm but then I wound up in the hospital because I drove myself into the ground so it brings back stressful memories... I usually stop at this point and listen to others...

Re: [agi] Re: Noise and Confirmation Bias

Well, if there is a definition of "lossy" that is consistent with "lossless" it certainly isn't in the vernacular nor in the canonical technical usage of those terms. What's you argot? On Thu, Oct 10, 2019 at 1:31 PM John Rose wrote: > On Thursday, October 10, 2019, at 2:26 PM, James Bowery

Re: [agi] Re: Noise and Confirmation Bias

On Thursday, October 10, 2019, at 2:26 PM, James Bowery wrote: > KC can be approximated *only* with lossless compression. Thanks for that valuable tidbit, but did you imply earlier that lossy and lossless are mutually exclusive? I'm not convinced of that but am not an expert even though I

Re: [agi] Re: Noise and Confirmation Bias

methinks ppl here know something about somethin i dont... -- Artificial General Intelligence List: AGI Permalink: https://agi.topicbox.com/groups/agi/T8c8ee84b385720a5-M3c8d59cda23127819a943f21 Delivery options:

Re: [agi] Re: Noise and Confirmation Bias

Yes, approximations of KC are critical to anyone serious about AGI, but keep in mind KC can be approximated *only* with lossless compression. On Thu, Oct 10, 2019 at 1:21 PM John Rose wrote: > Kolmogorov Complexity is estimable it helps. > *Artificial General Intelligence List

Re: [agi] Re: Noise and Confirmation Bias

That sounds like a world of thought there,  im not schooled in. But Ockhams razor means things to me,   in engineering, if its simpler it usually is better,  its like a sentence with a more concise and compact meaning.    then the rest of the inventions hit the razor, and they never see the

Re: [agi] Re: Noise and Confirmation Bias

Kolmogorov Complexity is estimable it helps. -- Artificial General Intelligence List: AGI Permalink: https://agi.topicbox.com/groups/agi/T8c8ee84b385720a5-M159f94abeab3cc9962f23249 Delivery options: https://agi.topicbox.com/groups/agi/subscription

Re: [agi] Re: Noise and Confirmation Bias

Confirmation bias is violation of the "necessity" in Ockham's Razor: Do not multiply entities beyond necessity. The choice to eliminate observations that you _think_ are "noise" or you _think_ have no implication for shaping models to be more predictive, is unjustifiable in the limit of Solomonff

Re: [agi] Re: Noise and Confirmation Bias

Are there any unbiased lossy compression algorithms? I would speculate that it's possible but practically assume there are none. -- Artificial General Intelligence List: AGI Permalink:

[agi] Finding a Simple Distribution In a Binary String

Unsupervised feature discovery of text that started with its bit string representation would need to discover octets were the first-order parse of such a bit string. This raises a question: What is the technique called that can discover that a binary string, for example:

Re: [agi] Re: Noise and Confirmation Bias

 Perfect noise should not be biased. -- Artificial General Intelligence List: AGI Permalink: https://agi.topicbox.com/groups/agi/T8c8ee84b385720a5-M13f71e3af146e33eeed0321e Delivery options: https://agi.topicbox.com/groups/agi/subscription

Re: [agi] Re: Noise and Confirmation Bias

 When a steady five value is bit streamed into a detector and and is subtracted from previous sampled, the values should be zero. BUT if look at it through a microscope and record them into a file and they bounce at around at a  two bit value and if there is a equal amount of recorded  a, b,

Re: [agi] Re: Noise and Confirmation Bias

I dont know what you mean by confirmation bias,  but lossy matching is definitely imo the most important thing a neural network does for people. even tho its actually slower, and exact lossless matching is quicker, starts off simpler to code, but ends up more complex in the end to get to

Re: [agi] Re: Noise and Confirmation Bias

Huffman coding\n for data compression.     nibble system https://www.techiedelight.com/huffman-coding/   |                                            /\              

Re: [agi] Re: Noise and Confirmation Bias

Explosive logorrhea, that's when you can't control your decompression. -- Artificial General Intelligence List: AGI Permalink: https://agi.topicbox.com/groups/agi/T8c8ee84b385720a5-M3baa707ada6df09d157b7fd7 Delivery options:

Re: [agi] Re: Speedrun continues

Pybol Cobon On Wed, Oct 9, 2019 at 5:24 PM wrote: > Name it something that stands for its ease of use. EasyPython > *Artificial General Intelligence List * > / AGI / see discussions + > participants

Re: [agi] Re: Noise and Confirmation Bias

Methinks the torrent responding to John Rose's self-contradictory question is an example of Explosive Logorrhea. On Thu, Oct 10, 2019 at 11:12 AM John Rose wrote: > ya there must be some existing hybrid out there > *Artificial General Intelligence List * > / AGI

Re: [agi] Made a corner detector

sorry S.R.  I actually didnt read what you posted before in a rushed excitement,   yes,  handling the "hand" is very important to me, too,  I will be trying to do it. -- Artificial General Intelligence List: AGI Permalink:

Re: [agi] Made a corner detector

we need success!! code is good if you can read it,  otherwise its as useless as any other kind of notation you havent learnt yet. -- Artificial General Intelligence List: AGI Permalink:

Re: [agi] Made a corner detector

Sorry, but it's just sad to be detached from success like that. So an AGI is not about code? On Thu, 10 Oct 2019 at 13:08, wrote: > The sentence is easier to understand than the code, code is shit, pseudo > code means more. > *Artificial General Intelligence List

Re: [agi] Made a corner detector

You could see it as a star constellation, or It could be the picture underneath the points. -- Artificial General Intelligence List: AGI Permalink: https://agi.topicbox.com/groups/agi/Tc5e7e8b165534b2f-M79bfa3e77d401cdd2d699a21 Delivery options:

Re: [agi] Made a corner detector

OK, on second reading, I do get it more. You are finding pictures, right, and a corner is a kind of picture? On Thu, 10 Oct 2019 at 13:08, wrote: > The sentence is easier to understand than the code, code is shit, pseudo > code means more. > *Artificial General Intelligence List

Re: [agi] Made a corner detector

non maximum suppressing. i mean, by suppression,  you also call it inhibition - if you want.  the corners fight for supremecy and only the strongest one wins. -- Artificial General Intelligence List: AGI Permalink:

Re: [agi] Made a corner detector

Its a pretty hacky one,   no-where near as good as a harris corner,  but its cheating its way better by some temporal stability i do over now and the last frame,  to knock off ones that arent solid enough reports. Without that its pretty bad. The corner detector I use is a nonthresholded sobel,

Re: [agi] Made a corner detector

What general approach are you using?  Is this a standard type of corner detector?  Something all-new?  In between? -- Artificial General Intelligence List: AGI Permalink: https://agi.topicbox.com/groups/agi/Tc5e7e8b165534b2f-M2859b9d07ccabfc9f9a153bc

Re: [agi] Re: Noise and Confirmation Bias

what about plastic morph loss,  which helps glue it all together,  and lossless is too rigid. -- Artificial General Intelligence List: AGI Permalink: https://agi.topicbox.com/groups/agi/T8c8ee84b385720a5-M295a1716370c71e6e67a7314 Delivery options:

Re: [agi] Re: Noise and Confirmation Bias

actually lossylossless is merely lossy with object recognition -- Artificial General Intelligence List: AGI Permalink: https://agi.topicbox.com/groups/agi/T8c8ee84b385720a5-M76c14dfbf2fea1b4fc5ab634 Delivery options:

Re: [agi] Re: Noise and Confirmation Bias

yes!! =D -- Artificial General Intelligence List: AGI Permalink: https://agi.topicbox.com/groups/agi/T8c8ee84b385720a5-Mf82f0f087eeb27d9ee334676 Delivery options: https://agi.topicbox.com/groups/agi/subscription

Re: [agi] Re: Noise and Confirmation Bias

I used to do that as a kid. -- Artificial General Intelligence List: AGI Permalink: https://agi.topicbox.com/groups/agi/T8c8ee84b385720a5-M81c2d333fcf7a4b79e2f7aaf Delivery options: https://agi.topicbox.com/groups/agi/subscription

Re: [agi] Re: Noise and Confirmation Bias

lossylossnessness is jpeg compressing a png. :) -- Artificial General Intelligence List: AGI Permalink: https://agi.topicbox.com/groups/agi/T8c8ee84b385720a5-M1e6a3bd6422563a4f86fefcc Delivery options: https://agi.topicbox.com/groups/agi/subscription

Re: [agi] Re: Noise and Confirmation Bias

Compressing a neutrosophic or fuzzy logic interpretation into boolean logic. That might be lossylosslessness. -- Artificial General Intelligence List: AGI Permalink: https://agi.topicbox.com/groups/agi/T8c8ee84b385720a5-Mb7d06ed09a35162270a8ea56 Delivery

Re: [agi] Re: Noise and Confirmation Bias

one unified key structure for every cell its database of never ending garbage. -- Artificial General Intelligence List: AGI Permalink: https://agi.topicbox.com/groups/agi/T8c8ee84b385720a5-Me97e5c05f3604d3e428b61ff Delivery options:

Re: [agi] International AI Issues

On Thursday, October 10, 2019, at 8:18 AM, Matt Mahoney wrote: > Backchannel uses TOR so it requires an internet connection. They use a modified TOR, but I have a relationship with them so I will avoid discussing... What do you think of QUIC which is rolling out:

Re: [agi] International AI Issues

Backchannel uses TOR so it requires an internet connection. Steve had an idea a few years ago for direct P2P intrrner between phones over WiFi, which the protocol supports but is not widely used because of a bug (now fixed) that caused the protocol to drastically slow down. The bug was a 32 bit

Re: [agi] Made a corner detector

put it to some nasty business. https://www.youtube.com/watch?v=IinqMCyeoqM=youtu.be -- Artificial General Intelligence List: AGI Permalink: https://agi.topicbox.com/groups/agi/Tc5e7e8b165534b2f-Me28bdfafa5c52067585d0f28 Delivery options:

Re: [agi] Made a corner detector

The sentence is easier to understand than the code,  code is shit,  pseudo code means more. -- Artificial General Intelligence List: AGI Permalink: https://agi.topicbox.com/groups/agi/Tc5e7e8b165534b2f-Ma76d16aaa1c681a2d89d6cc4 Delivery options:

Re: [agi] Made a corner detector

Hey > If u want the code or a nice flow chart, I can help out alot, I say if you dont tell another scientist something useful, you should get told stuff all yourself. Do I have to understand this sentence? :-) I was just asking for code... On Thu, 10 Oct 2019 at 12:11, wrote: >

Re: [agi] Re: Noise and Confirmation Bias

the problem with boolean logics is that they're boolean -- Artificial General Intelligence List: AGI Permalink: https://agi.topicbox.com/groups/agi/T8c8ee84b385720a5-M71663465485fdb4427d1700d Delivery options:

Re: [agi] Re: Noise and Confirmation Bias

then the whole chunk of data maps to one new symbol -- Artificial General Intelligence List: AGI Permalink: https://agi.topicbox.com/groups/agi/T8c8ee84b385720a5-M00bf9fc46c5f0711f6a8a9b5 Delivery options: https://agi.topicbox.com/groups/agi/subscription

Re: [agi] Re: Noise and Confirmation Bias

if its not obvious, theres a chance no-one found it, but it also could mean it actually IS impossible. -- Artificial General Intelligence List: AGI Permalink: https://agi.topicbox.com/groups/agi/T8c8ee84b385720a5-M812be927dee6aa639356dc29 Delivery options:

Re: [agi] Re: Noise and Confirmation Bias

How about resolving an alphabet or a language out of data where it is not obvious. -- Artificial General Intelligence List: AGI Permalink: https://agi.topicbox.com/groups/agi/T8c8ee84b385720a5-M59cd38581e06a08207daf90f Delivery options:

Re: [agi] Made a corner detector

https://www.youtube.com/watch?v=NoMQdN2CN0E=youtu.be cause of the uncanny detection computers can do if you get it right it can spot corners in extreme low light conditions,  because it can find corners on your cheeks.   But as a finished eye system probably only ends up modelling the

Re: [agi] Made a corner detector

A use case I have is recognizing hand gestures (e.g. number of fingers). Looks like this might help. Is there code? :) On Thu, 10 Oct 2019 at 10:43, wrote: > heaps more stable response using nearmatching of this and the last frame, > heaps more to fix tho... be back soon... >

Re: [agi] Made a corner detector

heaps more stable response using nearmatching of this and the last frame,    heaps more to fix tho...  be back soon... https://www.youtube.com/watch?v=m65zbBLLA54=youtu.be -- Artificial General Intelligence List: AGI Permalink: