Why PICS is the wrong approach

PICS is the wrong approach becuase it oversimplifies the ratings of
content, because it places the ratings made by the author in the payload
itself, and because third-party ratings systems are cut out of the loop
(effectively).

One computerish way to think of this is that the "binding" is too early. At
the time of distribution, say, I mark my work something with some PICS
label, based upon my best understanding of the PICS labels, ratings,
agencies, and laws. But once set, the "binding" has been made. Later
reviews or reviews by other entities cannot affect the binding, at least
not for this distributed instance.

And of course it is quite likely that things important to others in their
ratings are not as important to me. I might even ignore certain points, not
even seeing the need to point out things in the work. This is inevitable,
as there is no uniform view of truth, no uniform set of values and
priorities, and no hope there ever can be such a monistic view. Consider
the recent example of AOL's lists of banned words, even words in "harmless
situations" (e.g, the example someone cited of "tits" being banned, despite
being the name of a bird...would an animal-lovers Web page or posting with
"Tits and Asses!!!" prominently in the title be PICS labelled as obscene?
Some would surely think so.).

A much better solution is to let the unique ID block of an article--the
Usenet article ID, or some hash of the headers, whatever--be a pointer that
other ratings servies could then use to provide for their customers or
clients as a filtering mechanism. This would allow as many ratings services
to exist as clients would be willing to support.

Sure, there are _time delays_ in the evaluation process, as, for example,
the Catholic Index reviews Web pages and Usenet posts, but all evaluation
causes delay. This puts the burden on those proposing to filter content.

More importantly, the "payload" does not carry some particular set of
fairly-arbitrary PICS evluations. Binding by the censors instead of by the
originator, which is as it should be.

--Tim May


Just say "No" to "Big Brother Inside"
We got computers, we're tapping phone lines, I know that that ain't allowed.
---------:---------:---------:---------:---------:---------:---------:----
Timothy C. May              | Crypto Anarchy: encryption, digital money,
tcmay@got.net  408-728-0152 | anonymous networks, digital pseudonyms, zero
W.A.S.T.E.: Corralitos, CA  | knowledge, reputations, information markets,
Higher Power: 2^1398269     | black markets, collapse of governments.
"National borders aren't even speed bumps on the information superhighway."

Why TCM is wrong about PICS being wrong:

>PICS is the wrong approach becuase it oversimplifies the ratings of
>content, because it places the ratings made by the author in the payload
>itself, and because third-party ratings systems are cut out of the loop
>(effectively).

bzzzzzzt. please read about it. there are multiple protocols. some
of them allow third-party rating services. some of them support
ratings within pages. the standard is neutral.

>One computerish way to think of this is that the "binding" is too early. At
>the time of distribution, say, I mark my work something with some PICS
>label, based upon my best understanding of the PICS labels, ratings,
>agencies, and laws. But once set, the "binding" has been made. Later
>reviews or reviews by other entities cannot affect the binding, at least
>not for this distributed instance.

you have a good point, but PICS is about letting the net decide. it
supports both self-rated and third-party ratings. we will see whether
one eclipses the other in the long term. personally I suspect both
will coexist.

>And of course it is quite likely that things important to others in their
>ratings are not as important to me. I might even ignore certain points, not
>even seeing the need to point out things in the work. This is inevitable,
>as there is no uniform view of truth, no uniform set of values and
>priorities, and no hope there ever can be such a monistic view.

this is a ridiculous misunderstanding of the rating system concept.
the PICS standard expressly supports diversity by letting a thousand
rating services bloom, to borrow a phrase from your own book.  some
rating services may claim to be canonical, but you don't have to
believe them. there will be competition of rating services for a long
time into the future. this has already happened with all the filtering
software out there. 

also consider the new Firefly system that doesn't
actually have fixed ratings on objects, but in which ratings are 
determined dynamically based on your own personal ratings of pages.

 Consider
>the recent example of AOL's lists of banned words, even words in "harmless
>situations" (e.g, the example someone cited of "tits" being banned, despite
>being the name of a bird...would an animal-lovers Web page or posting with
>"Tits and Asses!!!" prominently in the title be PICS labelled as obscene?
>Some would surely think so.).

this would be an example of the most rudimentary and simplistic filtering
or rating service, which of course the market would generally ignore
in favor of more sophisticated alternative schemes.

>A much better solution is to let the unique ID block of an article--the
>Usenet article ID, or some hash of the headers, whatever--be a pointer that
>other ratings servies could then use to provide for their customers or
>clients as a filtering mechanism. This would allow as many ratings services
>to exist as clients would be willing to support.

that's exactly what PICS is about when you read about it more deeply.

>More importantly, the "payload" does not carry some particular set of
>fairly-arbitrary PICS evluations. Binding by the censors instead of by the
>originator, which is as it should be.

PICS supports both, as it was expressly designed to.

what Timmy is repeatedly failing to comprehend despite much
evidence staring him in the face is that ratings services are
going to be a very significant new information industry, if they
haven't already become one. there are now many different filtering
packages out there and the market is large for them, as has
been proven by *existing* sales. this industry will grow. yahoo
and many other indexing services are in fact implicitly
rating systems, because they utilize editorial discrimination in
deciding who to include and who to exclude.  they just don't say,
"this is rated yahoo approved" overtly.

(timmy is also upset that a massive new industry is growing without
his personal approval or anticipation. I will amuse myself by
counting the days until he does a flip in position and begins
to advocate rating system's efficacy while pretending his position
was never otherwise)

let a thousand rating systems bloom. PICS is about finding good
content as much as rejecting uninteresting content.

Vladimir Z. Nuri wrote:
> Why TCM is wrong about PICS being wrong:
> >PICS is the wrong approach becuase it oversimplifies the ratings of
> >content, because it places the ratings made by the author in the payload
> >itself, and because third-party ratings systems are cut out of the loop
> >(effectively).

> bzzzzzzt. please read about it. there are multiple protocols. some
> of them allow third-party rating services. some of them support
> ratings within pages. the standard is neutral.

> also consider the new Firefly system that doesn't
> actually have fixed ratings on objects, but in which ratings are
> determined dynamically based on your own personal ratings of pages.

If Firefly is an example of what PICS is or could become, the hell with
PICS.  Firefly encourages and rewards group behavior and suppresses
individuality.  Firefly would reward the discussion of the latest album
by a Columbia or Capitol artist, and discourage discussion of material
from independent (real independent) labels.  I know because I've been
there and spent quite a bit of time trying to get a rating.

[remainder snipped]

>If Firefly is an example of what PICS is or could become, the hell with
>PICS.  Firefly encourages and rewards group behavior and suppresses
>individuality.  Firefly would reward the discussion of the latest album
>by a Columbia or Capitol artist, and discourage discussion of material
>from independent (real independent) labels.  I know because I've been
>there and spent quite a bit of time trying to get a rating.
>

this is absolutely ridiculous. the rating system is designed to
be incredibly individualized. it uses sophisticated statistical
techniques to find the correlations in your UNIQUE ratings given
over a set of items with other people's ratings, and weighs
future ratings based on these correlations. it may be even
dealing with anti-correlations. in fact what you have, in effect,
is a system with *no* hardwired ratings. the ratings space is
different for *every*single*person* who uses the service.

"you know"? get a clue, please. these systems are not at all like
the record rating systems you get in stores. I sympathize with
your plight however and agree that the record labeling system
is a lame way to go about it. all the more reason to support things
like firefly and grouplens (another interesting system that may
have inspired firefly, do a yahoo search to find it). superior
rating systems will begin to flourish instead of inferior ones...

Vladimir Z. Nuri wrote:
> >If Firefly is an example of what PICS is or could become, the hell with
> >PICS.  Firefly encourages and rewards group behavior and suppresses
> >individuality.  Firefly would reward the discussion of the latest album
> >by a Columbia or Capitol artist, and discourage discussion of material
> >from independent (real independent) labels.  I know because I've been
> >there and spent quite a bit of time trying to get a rating.

> this is absolutely ridiculous. the rating system is designed to
> be incredibly individualized. it uses sophisticated statistical
> techniques to find the correlations in your UNIQUE ratings given
> over a set of items with other people's ratings, and weighs
> future ratings based on these correlations. it may be even
> dealing with anti-correlations. in fact what you have, in effect,
> is a system with *no* hardwired ratings. the ratings space is
> different for *every*single*person* who uses the service.

As I said, I spent considerable time there, and typed in considerable
data.  I used the help facilities, and I corresponded with the persons
who do support.  Firefly was not able to make a single correlation or
suggestion for me, which BTW, Alexander Chislenko acknowledged as a
weakness of the (current/old) Firefly system.  Perhaps you should ask him
what the problem is, if you have a curiosity about it.

[snip]