On Thursday, May 27, 2021, at 11:05 AM, James Bowery wrote:
> Something else that has occurred to me about lossless vs lossy aka AIT vs SDT:
>
On Thursday, May 27, 2021, at 11:05 AM, James Bowery wrote:
> The bit string to be losslessly compressed by AIT has to come from somewhere
> and that
On Thursday, May 27, 2021, at 5:52 AM, John Rose wrote:
> I would consider OCR (Optical Character Recognition) as a compression. The
> characters extracted are lossless and the noise and background is lossily
> compressed out.
Yes but... The compression of text in the Hutter Prize or LTBC is
Your response was submitted one minute after another submission by me
describing the evolution of embodied data input as a highly "lossy" channel
whose "utility function" reduces to physical replication. So, yes, our
sensory apparatuses are highly evolved to discard vast quantities of data
as
On Wed, May 26, 2021, 11:24 AM James Bowery wrote:
>
> The "noise" level, in AIT terms, can be thought of as a constant of
> integration and what we're interested in while searching the space of
> algorithms is differentiability of the loss function (ie: the number of
> bits in the executable
Something else that has occurred to me about lossless vs lossy aka AIT vs
SDT:
The bit string to be losslessly compressed by AIT has to come from
somewhere and that "somewhere" has to be, in some sense, "embodied" in
its environment so as to receive data from its environment. The structure
of
On Wednesday, May 26, 2021, at 11:23 AM, James Bowery wrote:
> In AIXI terms, the difference between lossless and lossy compression is the
> difference between AIT's and SDT's notion of "utility": The former being
> concerned with what "is" and the latter being concerned with what "ought" to
>
It's "impossible" that vision is any different from text Lossless Compression.
Patterns = better Lossless Compression. Lossy is just good fruit with some bad
side dish, Lossless Compression is all about looking at good fruit, things you
can return back from the dead by decompression. Vision
On Sun, May 23, 2021 at 9:47 PM Matt Mahoney
wrote:
> On Sun, May 23, 2021 at 1:57 PM James Bowery wrote:
> > On Sun, May 23, 2021 at 12:32 PM Matt Mahoney
> wrote:
> >>
> >> ...Data compression alone doesn't lead to AGI, but it does measure
> prediction in signals with a high signal to noise
(mine)
--
Artificial General Intelligence List: AGI
Permalink:
https://agi.topicbox.com/groups/agi/T95f11a183fb9b6e1-M779d63dbc12e7dc0c91e9c55
Delivery options: https://agi.topicbox.com/groups/agi/subscription
may not be so diverse tho
--
Artificial General Intelligence List: AGI
Permalink:
https://agi.topicbox.com/groups/agi/T95f11a183fb9b6e1-M38c1e5e86ba2b313b022c4b8
Delivery options: https://agi.topicbox.com/groups/agi/subscription
unless u use human made illustrations as a dataset - I actually have thousands
of human made drawings on my computer lol.
--
Artificial General Intelligence List: AGI
Permalink:
> mages are not made by brains
(I mean, OUR images aren't made by brains, they are only snapshotted by humans
and that's it, or the items are organized by us, or made by us in
factories...but mostly not by us exactly...) (brains do generate images but um
I CAN'T SHOW YOU MINE IN MY BRAIN lol
I only like text prediction because using human made patterns is the best,
images are not made by brains, so there may be less patterns from nature. But I
will not limit myself to text one day. And a better compressor will compress
images better, I see no reason to not do such a contest.
On Sun, May 23, 2021 at 1:57 PM James Bowery wrote:
> On Sun, May 23, 2021 at 12:32 PM Matt Mahoney wrote:
>>
>> ...Data compression alone doesn't lead to AGI, but it does measure
>> prediction in signals with a high signal to noise ratio, like text. It's
>> less useful for vision and
On Sun, May 23, 2021 at 12:32 PM Matt Mahoney
wrote:
> ...Data compression alone doesn't lead to AGI, but it does measure
> prediction in signals with a high signal to noise ratio, like text. It's
> less useful for vision and robotics.
>
Seems to me if a vision system can transform a 2D array
On Sunday, May 23, 2021, at 1:30 PM, Matt Mahoney wrote:
> It's less useful for vision and robotics.
Data compression is useful for vision, it is just a couple of extra pattern
finding mechanisms that make it work on images. We saw that GPT tech on
openAI.com works wonders in DALL-E, the
On Sat, May 22, 2021, 12:25 PM Alan Grimes via AGI
wrote:
> immortal.discover...@gmail.com wrote:
> > I am on this list, I am completely sane
>
> Ha! You think text compression is a useful avenue for reaching AGI! =P
My phone tells me what song is playing on the radio. I can point it at a
sign
On Saturday, May 22, 2021, at 12:47 PM, Alan Grimes wrote:
> binding
DALL-E on openAI.com does this one, you can see how advanced DALL-E is.
On Saturday, May 22, 2021, at 12:47 PM, Alan Grimes wrote:
> we need tothe AGI to be capable of the full range of human activities
including perception,
immortal.discover...@gmail.com wrote:
> By programming just a dozen pattern finding mechanisms, you build a
> general purpose base that can build from there up on its own and make
> thousands of more precise rules.
Which is why you are insane!
You can't make predictions that are either useful or
It's not a compressor, it's a predictor of the next letter of a sentence. It so
happens that, being able to discover the rest of a sentence that you never seen
before, allows you to step through a text file and compress it by not having to
store much since you can discover the rest (by using
immortal.discover...@gmail.com wrote:
> I am on this list, I am completely sane
Ha! You think text compression is a useful avenue for reaching AGI! =P
--
The vaccine is a LIE.
#EggCrisis
White is the new Kulak.
Powers are not rights.
--
Artificial
I am on this list, I am completely sane and here and see the world for as it is
(we are in hell cooking and I am upset about this ugly universe, there's an
animal in pain somewhere as we speak), and I am serious about AGI (I made a
whole complete AGI guide and am making the algorithm, I can
22 matches
Mail list logo