zenon/trustworthy-ai
1
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity

Change em-dash to ---

Browse Source
This commit is contained in:
Tobias Eidelpes 2022-01-05 15:13:32 +01:00
parent b07dc3d676
commit f5cf8fcd4e
1 changed files with 19 additions and 19 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

38
trustworthy-ai.tex
View File

@ -50,14 +50,14 @@ environment~\cite{russellArtificialIntelligenceModern2021}.
While the possibilities of AI are seemingly endless, the public is slowly but While the possibilities of AI are seemingly endless, the public is slowly but
steadily learning about its limitations. These limitations manifest themselves steadily learning about its limitations. These limitations manifest themselves
in areas such as autonomous driving and medicine, for example. These are fields in areas such as autonomous driving and medicine, for example. These are fields
where AI can have a direct—potentially life-changing—impact on people's lives. A where AI can have a direct---potentially life-changing---impact on people's
self-driving car operates on roads where accidents can happen at any time. lives. A self-driving car operates on roads where accidents can happen at any
Decisions made by the car before, during and after the accident can result in time. Decisions made by the car before, during and after the accident can result
severe consequences for all participants. In medicine, AIs are increasingly used in severe consequences for all participants. In medicine, AIs are increasingly
to drive human decision-making. The more critical the proper use and functioning used to drive human decision-making. The more critical the proper use and
of AI is, the more trust in its architecture and results is required. Trust, functioning of AI is, the more trust in its architecture and results is
however, is not easily defined, especially in relation to artificial required. Trust, however, is not easily defined, especially in relation to
intelligence. artificial intelligence.
This work will explore the following question: \emph{Can artificial intelligence This work will explore the following question: \emph{Can artificial intelligence
be trustworthy, and if so, how?} To be able to discuss this question, trust has be trustworthy, and if so, how?} To be able to discuss this question, trust has
@ -95,8 +95,8 @@ that the trustee does not violate our \emph{social agreement} by acting against
our interests. Often times we are not able to confirm that the trustee has our interests. Often times we are not able to confirm that the trustee has
indeed done his/her job. Sometimes we will only find out later that what did indeed done his/her job. Sometimes we will only find out later that what did
happen was not in line with our own interests. Trust is therefore also always a happen was not in line with our own interests. Trust is therefore also always a
function of time. Previously entrusted people can—depending on their track function of time. Previously entrusted people can---depending on their track
record—either continue to be trusted or lose trust. record---either continue to be trusted or lose trust.
We do not only trust certain people to act on our behalf, we can also place We do not only trust certain people to act on our behalf, we can also place
trust in things rather than people. Every technical device or gadget receives trust in things rather than people. Every technical device or gadget receives
@ -122,8 +122,8 @@ because of their existing relationship.
Due to the different dimensions of trust and its inherent complexity in Due to the different dimensions of trust and its inherent complexity in
different contexts, frameworks for trust are an active field of research. One different contexts, frameworks for trust are an active field of research. One
such framework—proposed by \textcite{ferrarioAIWeTrust2020}—will be discussed in such framework---proposed by \textcite{ferrarioAIWeTrust2020}---will be
the following sections. discussed in the following sections.
\subsection{Incremental Model of Trust} \subsection{Incremental Model of Trust}
@ -243,8 +243,8 @@ the model receives from its beneficiaries. One such example may be Netflix'
movie recommendation system that receives which type of movies certain users are movie recommendation system that receives which type of movies certain users are
interested in. A malicious user could therefore attack the recommendation engine interested in. A malicious user could therefore attack the recommendation engine
by supplying wrong inputs. \emph{Evasion attacks} consist of alterations which by supplying wrong inputs. \emph{Evasion attacks} consist of alterations which
are made to the training samples in such a way that these alternations—while are made to the training samples in such a way that these alternations---while
generally invisible to the human eye—mislead the algorithm. generally invisible to the human eye---mislead the algorithm.
\emph{White-box attacks} allow an attacker to clearly see all parameters and all \emph{White-box attacks} allow an attacker to clearly see all parameters and all
functions of a model. \emph{Black-box attackers}, on the other hand, can only functions of a model. \emph{Black-box attackers}, on the other hand, can only
@ -271,11 +271,11 @@ trade-offs is therefore critical for real-world applications.
Non-discrimination and fairness are two important properties of any artificial Non-discrimination and fairness are two important properties of any artificial
intelligence system. If one or both of them are violated, trust in the system intelligence system. If one or both of them are violated, trust in the system
erodes quickly. Often researchers only find out about a system's discriminatory erodes quickly. Often researchers only find out about a system's discriminatory
behavior when the system has been in place for a long time. In other cases—such behavior when the system has been in place for a long time. In other
as with the chat bot Tay from Microsoft Research, for example—the problems cases---such as with the chat bot Tay from Microsoft Research, for example---the
become immediately apparent once the algorithm is live. Countless other models problems become immediately apparent once the algorithm is live. Countless other
have been shown to be biased on multiple fronts: the US' recidivism prediction models have been shown to be biased on multiple fronts: the US' recidivism
software \textsc{COMPAS} is biased against black people prediction software \textsc{COMPAS} is biased against black people
\cite{angwinMachineBias2016}, camera software for blink detection is biased \cite{angwinMachineBias2016}, camera software for blink detection is biased
against Asian eyes \cite{roseFaceDetectionCamerasGlitches2010} and gender-based against Asian eyes \cite{roseFaceDetectionCamerasGlitches2010} and gender-based
discrimination in the placement of career advertisements discrimination in the placement of career advertisements