Re: Opening the AI Black Box

The Government of the Netherlands also uses algorithmic decision making systems for arriving at court verdicts in cases where the court cases do not really require a magistrate to make a decision.
These systems are however creating a growing resistance, and require a legal redress system where a human reviews AI errors made.
Sounds counterproductive, doesn't it? Milton Ponson
GSM: +297 747 8280
PO Box 1154, Oranjestad
Aruba, Dutch Caribbean
Project Paradigm: Bringing the ICT tools for sustainable development to all stakeholders worldwide through collaborative research on applied mathematics, advanced modeling, software and standards development 

    On Wednesday, September 26, 2018 8:25 AM, Owen Ambur <Owen.Ambur@verizon.net> wrote:
 

 #yiv1387142236 #yiv1387142236 -- _filtered #yiv1387142236 {panose-1:2 4 5 3 5 4 6 3 2 4;} _filtered #yiv1387142236 {font-family:Calibri;panose-1:2 15 5 2 2 2 4 3 2 4;}#yiv1387142236 #yiv1387142236 p.yiv1387142236MsoNormal, #yiv1387142236 li.yiv1387142236MsoNormal, #yiv1387142236 div.yiv1387142236MsoNormal {margin:0in;margin-bottom:.0001pt;font-size:11.0pt;font-family:sans-serif;}#yiv1387142236 a:link, #yiv1387142236 span.yiv1387142236MsoHyperlink {color:blue;text-decoration:underline;}#yiv1387142236 a:visited, #yiv1387142236 span.yiv1387142236MsoHyperlinkFollowed {color:purple;text-decoration:underline;}#yiv1387142236 p.yiv1387142236msonormal0, #yiv1387142236 li.yiv1387142236msonormal0, #yiv1387142236 div.yiv1387142236msonormal0 {margin-right:0in;margin-left:0in;font-size:11.0pt;font-family:sans-serif;}#yiv1387142236 span.yiv1387142236EmailStyle18 {font-family:sans-serif;color:windowtext;}#yiv1387142236 .yiv1387142236MsoChpDefault {font-family:sans-serif;} _filtered #yiv1387142236 {margin:1.0in 1.0in 1.0in 1.0in;}#yiv1387142236 div.yiv1387142236WordSection1 {}#yiv1387142236 The U.S. House Committee on Oversight and Government Reform Subcommittee on Information Technology has issued a report entitled "Rise of the Machines: Artificial Intelligence and its Growing Impact on U.S. Policy."   One of the report's conclusions is that "a widely agreed upon standard for measuring the safety and security of AI products and applications should precede any new regulations." https://oversight.house.gov/wp-content/uploads/2018/09/AI-White-Paper-.pdf   Chairman Hurd emphasized the need for a national strategy and both he and ranking minority member Kelly referenced the importance of open data.  Two logical conclusions that might be drawn are:   1) the AI strategy itself should be published in an open, standard, machine-readable format like Strategy Markup Language (StratML) Part 1 (ISO 17469-1) and    2) AI product and application safety and security metrics should be reported in a format like StratML Part 2, Performance Plans and Reports.  https://fcw.com/articles/2018/09/25/hurd-kelly-ai-machines.aspx  Owen AmburChair, StratML Working GroupCo-Chair Emeritus, xml.gov CoPWebmaster, FIRMProfile on LinkedIn | Personal Home Page    From: Paola Di Maio <paola.dimaio@gmail.com> 
Sent: Wednesday, September 26, 2018 6:51 AM
To: public-aikr@w3.org
Subject: Opening the AI Black Box    Greetings, back from Japan,  I was quiet because a) both my computer crashed and b) I relocatedto another country in Asia, I am now in Tawan  Japan was a research trip in part relevant to the interest area of this group  Much AI needs more high level thinkingIt seems that most universities are teaching ML algos but not KR  I think the work of our group is much needed  Relevant report via twitter  CHALLENGING GOVERNMENT USE OF ALGORITHMIC DECISION SYSTEMS  https://ainowinstitute.org/litigatingalgorithms.pdf  I ll add this to the resources and may perform some key word extractionto pick out relevant concepts and terms  p  Dr Paola Di MaioNCKUA bit about me