Communication en Anti-Pattern: Insensitive Bot Dialogue <span>Anti-Pattern: Insensitive Bot Dialogue</span> <div> <div>Application</div> <div><a href="/taxonomy/term/6" hreflang="en">Communication</a></div> </div> <span><span>leighbryant</span></span> <span>Wed, 08/21/2019 - 18:26</span> <div> <div>Topic</div> <div><a href="/taxonomy/term/46" hreflang="en">Fairness &amp; Inclusiveness</a></div> </div> <div><p><strong>Problem:</strong></p> <p>The user does not want to be insulted, offended, triggered, or otherwise unsettled by a chatbot.</p> <figure><img alt="Side-by-side screenshots showing a chatbot responding to user queries with insensitive dialogue." data-entity-type="file" data-entity-uuid="ae40b0a0-ce58-4072-8f18-ef5722aa6da2" src="/sites/default/files/content-images/Insensitive_Bot_Dialogue-pandorabots_0.png" /> <figcaption>In both examples above, the chatbot response is likely to be interpreted by at least a few people as offensive or demeaning.</figcaption> </figure> <p><strong>Anti-pattern response:</strong></p> <p>The bot employs insensitive dialogue that can insult, offend, cause the user to be triggered, or otherwise unsettled.&nbsp;</p> <p>&nbsp;</p> <p><strong>Discussion:</strong></p> <p>Whether through inappropriate jokes, ugly associations, dubious assumptions, or just thoughtlessness in choice of words, a pleasant experience with a bot can turn into an awful one. In general, the cause is a designer who assumes all users share their worldview, their prejudices, and their sense of humour.</p> <p>Does this mean that bot dialogue should be bland and unremarkable in order to be safely inoffensive? Absolutely not. Brilliant copy can be witty, sharp, and inclusive. It does mean, however, that copywriting such dialogue should not be seen as finessing an otherwise effective product. As far as the user is concerned, the dialogue is the product.<br /> &nbsp;</p> </div> Wed, 21 Aug 2019 18:26:35 +0000 leighbryant 146 at Apologies for Inaccuracy <span>Apologies for Inaccuracy </span> <div> <div>Application</div> <div><a href="/taxonomy/term/6" hreflang="en">Communication</a></div> </div> <span><span>leighbryant</span></span> <span>Wed, 08/21/2019 - 17:17</span> <div> <div>Topic</div> <div><a href="/taxonomy/term/46" hreflang="en">Fairness &amp; Inclusiveness</a></div> </div> <div><p><strong>Problem:</strong></p> <p>The AI in a system may generate inaccurate results that cause confusion, may be offensive, or otherwise unsettle the user. The user would like to know that this is a possibility and feel comfortable with the system regardless.</p> <figure><img alt="Screenshot of an image of a man with a potentially inaccurate age guess attached to it, with a close up of an apology for the AI's possible inaccuracies. Example taken from" data-entity-type="file" data-entity-uuid="61ea0725-48f3-4490-ab9d-dd97d300dc16" src="/sites/default/files/content-images/Apologies_for_Inaccuracy-how-old-net-.png" /> <figcaption>Microsoft's "How Old" AI-powered software guesses the age of the individual featured, but also includes an apology disclaimer for inaccurate guesses beneath the image.</figcaption> </figure> <p>&nbsp;</p> <p><strong>Solution:</strong></p> <p>The system apologizes for any possible inaccuracy that may be present, especially in regards to sensitive topics that could cause harm or offence.</p> <p>&nbsp;</p> <p><strong>Discussion:</strong></p> <p>While it is important for systems to admit to their fallibility (via <a href="/patterns/21/setting-expectations-acknowledging-limitations">Setting Expectations &amp; Acknowledging Limitations</a>, for example), there is an additional requirement to address how this fallibility may affect the user. In reality, what we call "the user" is a million unique individuals each of whom want to be treated with dignity and respect, not just in this interaction but in every facet of everyday life. We need to use different lenses, with more expansive understandings of who "the user" is in terms of social, psychological, physical, and ideological variables that describe any one embodied individual that goes beyond standard task completion assessments. Apologizing for potentially harmful interpretations acknowledges the humanity of the individuals.</p> </div> Wed, 21 Aug 2019 17:17:43 +0000 leighbryant 126 at Risk Alert <span>Risk Alert</span> <div> <div>Application</div> <div><a href="/taxonomy/term/6" hreflang="en">Communication</a></div> </div> <span><span>leighbryant</span></span> <span>Wed, 08/21/2019 - 16:04</span> <div> <div>Topic</div> <div><a href="/taxonomy/term/36" hreflang="en">Transparency &amp; Trust</a></div> </div> <div><p><strong>Problem:</strong></p> <p>Users want to know how much trust to place in the system's intelligence, especially in critical applications such as the diagnosis of health issues, where there can be real world implications of following bad advice.</p> <figure><img alt="Example of reporting risk on Ada, a health app, about the limitations of the diagnosis functions of the app" data-entity-type="file" data-entity-uuid="0e88987d-f9ee-45ec-ae98-2fd3514caa03" src="/sites/default/files/content-images/Risk_Alert-ada_0.png" /> <figcaption>A medical app alerts users to the importance of seeking additional input from a professional.</figcaption> </figure> <p><strong>Solution:</strong></p> <p>In combination with <a href="/patterns/21/setting-expectations-acknowledging-limitations">Setting Expectations &amp; Acknowledging Limitations</a>, the system alerts the users of the risks&nbsp;they may encounter in following its advice, and how to mitigate that risk (i.e. to seek further advice from an expert before proceeding with any action).</p> <p>&nbsp;</p> <p><strong>Discussion:</strong></p> <p>While many faulty AI-driven predictions will result in little more than an ill-fitting dress or unliked movie, some are far more impactful. Great care should be taken in addressing the risks to the user, both because there can be potential legal and commercial implications, and because there is a moral imperative to reduce potential harm where possible.<br /> &nbsp;</p> </div> Wed, 21 Aug 2019 16:04:55 +0000 leighbryant 26 at Setting Expectations & Acknowledging Limitations <span>Setting Expectations &amp; Acknowledging Limitations </span> <div> <div>Application</div> <div><a href="/taxonomy/term/6" hreflang="en">Communication</a></div> </div> <span><span>leighbryant</span></span> <span>Wed, 08/21/2019 - 15:56</span> <div> <div>Topic</div> <div><a href="/taxonomy/term/36" hreflang="en">Transparency &amp; Trust</a></div> </div> <div><p><strong>Problem:</strong></p> <p>Users want to know what the AI being used in the application is actually capable of.</p> <figure><img alt="Example of setting user expectations on Ada by issuing a warning about the limitations of diagnostics in this health application" data-entity-type="file" data-entity-uuid="c815d39e-9bdb-4ffb-a356-06acb4a04bc7" src="/sites/default/files/content-images/Setting_Expectations-ada.png" /> <figcaption>By clarifying the parameters of the outputs the AI application is providing, the system sets clear expectations for the users in the quality of the results.</figcaption> </figure> <p><strong>Solution:</strong></p> <p>Before initiating a calculation or displaying results, the system tells the user what the AI can calculate, what the limits are in terms of accuracy in the output, which outputs are unlikely, and what is impossible to achieve that the user might otherwise want.</p> <p>&nbsp;</p> <p><strong>Discussion:</strong></p> <p>Trust is built on transparency, which necessarily includes communicating the limitations of the AI. Therefore this is partly about managing the user’s expectations throughout the process, but also about setting an appropriate tone of voice and brand position, one based on honesty and humility.<br /> &nbsp;</p> </div> Wed, 21 Aug 2019 15:56:19 +0000 leighbryant 21 at Explanation of Algorithm <span>Explanation of Algorithm</span> <div> <div>Application</div> <div><a href="/taxonomy/term/6" hreflang="en">Communication</a></div> </div> <span><span>leighbryant</span></span> <span>Wed, 08/21/2019 - 15:39</span> <div> <div>Topic</div> <div><a href="/taxonomy/term/36" hreflang="en">Transparency &amp; Trust</a></div> </div> <div><p dir="ltr"><strong>Problem:</strong></p> <p dir="ltr">The user&nbsp;wants to know, in general terms, how the AI is operating: what it takes as inputs, what calculations it runs, and what it outputs.</p> <figure><img alt="An image of a person created by an algorithm with an attached explanation of how the technology behind the image creation works" data-entity-type="file" data-entity-uuid="68f58b63-68ab-4c73-b2cd-02fb6a383502" src="/sites/default/files/content-images/explanation-of-algorithm-thispersondoesnotexist-cropped_0.png" /> <figcaption>A GAN-created image with a pop-out text box providing the source of the image as well as links explaining how this image was created.</figcaption> </figure> <p dir="ltr"><strong>Solution:</strong></p> <p dir="ltr">Either within the&nbsp;app or as a separate content area accessed via a link, there should be a concise but complete description of the AI's workings. It should be easily understood by lay readers and not&nbsp;obscure&nbsp;details through vague language, highly technical jargon, or&nbsp;misleading metaphors.</p> <p dir="ltr"><strong>Discussion:</strong></p> <p dir="ltr">The craving to understand how things work, while not manifested in every user, is fundamental to the human condition. In some cases, users want to understand so they can determine how to drive the result they’d like to see from the algorithm. Whatever the motivation for the user, providing an explanation is more than an exercise in building trust (though it is that, too)— it can also be an opportunity to delight by satisfying a recreational curiosity or offering the user additional control.&nbsp;</p> <p dir="ltr"><strong>More Examples:</strong></p> <figure><img alt="Screenshot of Microsoft's caption bot, with explanation of how the technology works" data-entity-type="file" data-entity-uuid="a446a001-b3f1-4a71-b02a-a7ddf6de373f" src="/sites/default/files/content-images/explanation-of-algorithm-captionbot.png" /> <figcaption>Microsoft's provides a wealth of information about how it works behind the scenes.</figcaption> </figure> <figure><img alt="Side-by-side screenshots of Bing and How Old Are You, both of which include explanations behind the technology" data-entity-type="file" data-entity-uuid="9d52d7cf-16a1-43c4-a733-6edd60e07e47" src="/sites/default/files/content-images/explanation-of-algorithm-bing%20and%20how%20old.png" /> <figcaption>More image recognition-based AI applications provide users with additional information on how they work.</figcaption> </figure> </div> Wed, 21 Aug 2019 15:39:36 +0000 leighbryant 16 at