Re: Bias & Perception

Thank you!
Please note that in AI terms, the question may be reversed

how does perception determine/cause bias?
is our cognitive bias influenced by our imperfect perceptual apparatus?

much to be worked on :-)


On Wed, Apr 22, 2020 at 9:41 AM Owen Ambur <Owen.Ambur@verizon.net> wrote:

> Here's Google's top hit on "how does bias affect perception":
> https://catalogofbias.org/biases/perception-bias/
>
> BTW, this exchange prompted me to convert to StratML format the Perception
> Institute's about statement, at
> https://stratml.us/drybridge/index.htm#PRCPTN  I wonder if they've given
> any thought to engaging with AI/ML agent developers or vice versa.
> Although their about statement makes no reference to "artificial" or
> "intelligence," I suspect they would have valuable expertise to lend to the
> cause of less biased algorithms.
>
> Owen
> On 4/21/2020 8:00 PM, Paola Di Maio wrote:
>
> Bias can be very complex and not well organised imho
> (work to be done)
> looks like what you point to is perceptual.
>
> On Wed, Apr 22, 2020 at 3:56 AM Owen Ambur <Owen.Ambur@verizon.net> wrote:
>
>> I haven't checked CEBM's catalog of biases
>> <https://stratml.us/drybridge/index.htm#COB> to see if it includes this
>> one -- https://rationalwiki.org/wiki/Style_over_substance -- but it
>> seems highly relevant to the work of the AIKR CG.
>>
>> Wikipedia's listing of cognitive biases --
>> https://en.wikipedia.org/wiki/List_of_cognitive_biases -- doesn't seem
>> to reference it directly.  However, attractiveness is referenced in these
>> biases:
>>
>> https://en.wikipedia.org/wiki/Cheerleader_effect
>>
>> https://en.wikipedia.org/wiki/Halo_effect
>>
>> I also discovered a separate article on bias, which includes this one:
>> https://en.wikipedia.org/wiki/Bias#Lookism
>>
>> Owen
>>
>> -------- Forwarded Message --------
>> Subject: Re: IPTC draft credibility guidelines released for feedback
>> Date: Tue, 21 Apr 2020 15:19:22 -0400
>> From: Owen Ambur <Owen.Ambur@verizon.net> <Owen.Ambur@verizon.net>
>> To: public-credibility@w3.org
>>
>> This is very good news, Brendan.
>>
>> The NewsCode Scheme is now available in StratML Part 1, Strategic Plan,
>> format at https://stratml.us/drybridge/index.htm#NCS
>>
>> Here are some comments, for whatever they may be worth:
>>
>>    1. While we must deal with reality as it currently exists, we should
>>    also pursue continuous improvement.
>>    2. We already have far too much "policy" in narrative format and far
>>    too few actual performance plans and reports, in open, standard,
>>    machine-readable format.
>>    3. While "short cuts" (like stories) are essential in the routine of
>>    everyday life, they exclude information (i.e., reliable data) that may be
>>    critical for consideration when risks may be high.
>>    4. HTML is a shortcut enabling the presentation of data.
>>    https://stratml.us/references/FlashyVIntelligentWeb.pdf
>>    5. The incumbents whose business cases and competitive advantages are
>>    based upon legacy data formats and the inefficiency of others should not be
>>    allowed to stand in the way of innovation and progress.
>>    6. We should aim for more mature business processes.
>>    https://en.wikipedia.org/wiki/Machine-readable_document
>>
>> I take this reference as further confirmation of my bias toward the
>> importance of the character of the content versus the style of the
>> presentation:  https://rationalwiki.org/wiki/Style_over_substance
>>
>> I wonder if Miki paid Google to make this their top hit on "bias toward
>> style versus substance":
>> https://www.megumimiki.com/blog/bias-towards-style-over-substance-is-keeping-your-real-talent-hidden
>>
>> Surely, there must be an evolutionary basis for our often irrational
>> attraction to attractiveness, commonly to the exclusion of factors more
>> relevant to the achievement of our objectives.
>> https://www.linkedin.com/pulse/artificial-ignorance-owen-ambur/
>>
>> See, for example,
>> https://www.sciencedirect.com/science/article/abs/pii/0162309595000682
>>
>> With reference to credibility "signals," this reference also uses that
>> term:
>> https://www.researchgate.net/publication/312719482_Evolutionary_Basis_of_Attraction
>>
>> Just some thoughts ... for whatever they might be worth.
>>
>> Owen
>> On 4/21/2020 5:06 AM, Brendan Quinn wrote:
>>
>> Hi Sandro and all,
>>
>> I have something to share: we've released the first public draft of our
>> "Expressing Trust and Credibility in IPTC Standards" document, as discussed
>> in a CredWeb call back in November.
>>
>> Here's our news item about it:
>>
>> https://iptc.org/news/public-draft-for-comment-expressing-trust-and-credibility-information-in-iptc-standards/
>>
>> All comments and feedback are gratefully accepted!
>>
>> Best regards,
>>
>> Brendan.
>>
>> On Mon, 20 Apr 2020 at 23:49, Sandro Hawke <sandro@w3.org> wrote:
>>
>>> Let's skip this week.
>>>
>>> Stay safe, and feel free to send the group email about interesting &
>>> relevant topics.
>>>
>>>        -- Sandro
>>>
>>>
>>>