Category Archives: Enterprise

Bots, Part One: Of bots, and bleeps, and other things

First published at the UMIT Newsletter as “Bots, Part One: Of bots, and bleeps, and other things.”


Screenshot of the ELIZA Talking website, showing an old VT100 digivox terminal with keyboard, microphone, speaker, and vox (voice) dial button. There is text on the monitor which provides a description of the project ("ELIZA is a mock Rogerian psychotherapist), credits, and instructions for use.

E.L.I.Z.A. Talking (www.masswerk.at/eliza/)

Last December, I was turned on to the importance of bots while attending the Mayo Clinic Social Media Network (MCSMN) Annual Meeting in Arizona. Since then, I’ve been digging into the topic, trying to learn more, and hoping to get a bot implemented on our departmental website, but there is just so much to talk about! We decided to break this into a few parts, beginning with defining terms and a snippet of history — what’s a bot?

Bots, as in…robots? Not exactly

My first question was, “What do they mean by ‘bots’?” No, not robots (even though we have several wonderful robotics initiatives going here on campus), NPCs in games, spiders, spambots, malware bots, botnets, zombie computers, or any other geeky gore. The folk at MCSMN were focusing solely on chatbots, also known as chatterbots, website chatbots, AI bots, intelligent agents, intelligent bots, talkbots, Twitterbots, or messenger bots on Facebook, or even just broadly as social bots. According to Chatbots.org, there are over 160 terms describing the idea of a chatbot!

I am old enough to think of these generally as AI conversation bots, like the original ELIZA, which I marveled at the first time I encountered it. We’ve come a long, LONG way since then…in some ways. Eliza is available in a javascript clone of the old terminal interface, and has a new version that will talk to you, but…in other ways, we still have a long way to go.

Bots then and now

Bots have a fascinating back story, with mind-blowing applications going on right now. They are on web sites and Facebook pages, in apps, and are used for everything from marketing and jobs to social justice and health interventions. New examples appear almost every day. Sometimes a bot is fairly simple, more of a script to automate tasks, while others edge into the areas related to artificial intelligence.

About the background of AI bots, there is some justification for saying they go back to Ancient Greece, which was a surprise to me. More recently, they took off again in the 1960s with bots like Eliza, and the 1970s with MYCIN, finally reaching the general public in the 1990s in games, and now we all know about Siri and Google Assistant, right? Futurism has an infographic on bots history, and it’s a topic that has really been growing.

In recent months, chatbots have been profiled in a variety of major media outlets, include Ad Week, CIO, CNBC, Forbes, TechCrunch, The Next Web, and Wired. Believe it or not, there are even a few magazines devoted to chatbots, including Chatbots Magazine, Chatbot News Daily, Chatbots Journal, and probably others I haven’t found. The short message is that if you haven’t been paying attention to bots for your websites or apps, now may be the time to change that.

Here at the University of Michigan, there are a lot of people studying bots and using bots to make interesting things, but you’ll hear more about that next time, in “Things to Do With Bots.”

Advertisements

Social Media For Exclusion

First posted (with a few small editorial changes) at the Mayo Clinic Center for Social Media Blog: http://socialmedia.mayoclinic.org/discussion/social-media-for-exclusion/


Pic of the day - Sweetness

We praise the power of social media to bring people together, to engage people around topics and events of shared interest, to include. Too often we don’t stop to think that every new tool with new ways to include people is probably excluding someone else in a new way.

The accessibility of social media is rarely discussed outside of communities of persons with disabilities. When it is discussed, the people who most need to hear it (us?) may not be in the audience. When we aren’t, the work we do to try to include and engage may actually serve to accomplish the opposite.

VIDEO

The obvious example, one you probably already know about, is when you share video or use streaming tools, but lack captions. It isn’t always possible to pay for live captioning, as much as we might want to, but there are tips and tricks to get around that with low-cost options. In Second Life, the Virtual Ability community has, as a standard of practice, a roster of volunteers who offer to type into live chat a synopsis of what the speaker is saying, and who describe images on the slides for the blind. Because some conversations there are held in typed chat, they also have someone sighted volunteer to read aloud the typed chat, identifying the speakers. Why couldn’t we working in healthcare media implement something like this for streamed events?

TWITTER

Did you know about Easy Chirp? It’s a Twitter browser designed for people who are blind or visually impaired, but it can be useful for others as well. Just as with curb cuts, tools to make social media more accessible for marginalized and excluded audiences can make life easier for people who are nominally able-bodied. Accessible browsers can require less bandwidth and be more robust when you are in that crowded meeting room with a hundred people sharing wireless, just as an example.

Do you tweet using one of those marvelous fancy tools for the pros that allow you to track multiple channels at the same time? Often those will place the relevant hashtag for the channel at the beginning of each tweet. I wager you weren’t aware that this makes life harder for those with visual or cognitive challenges, because it makes it harder for them to sort out the important part of the message. Recommended practice is to place hashtags at the END of your tweets.

Also, for those hashtags which smoosh together parts of different words? Screenreaders struggle to figure out how to read those when they are all lower case or all upper case. I have a bad habit of not mixing the capitalization because it is faster when I am typing, but for screenreader, there is a world of difference between #mayoclinic and #MayoClinic. Now that I know, I am trying to be better about it, because I was excluding people when I meant to include them.

MORE

These are just a few examples, low-hanging fruit. There is a lot more to know. Luckily, there is an upcoming free webinar on social media accessibility from the ADA National Network, if you are interested in learning more. If you are here on campus, we are meeting in the Hatcher Gallery Lab to watch this together.

October 20, 2015
2:00-3:30 p.m.
Social Media and Accessibility

#WHDemoDay and #ADAinitiative — Oh, the Irony


Welcome to Demo Day at the White House! (Megan Smith, the First US Chief Technology Officer) https://www.youtube.com/watch?v=PxGrDsuwCFk

“It’s a tradition in the tech community to show off amazing things that people have built. … All Americans do this. All American are capable of this. And it’s a big part of our future, and it’s always been a big part of our past.”

Yesterday was a landmark day in diversity and inclusion.

Yesterday saw the first ever White House Demo Day (#WHDemoDay), for women and minority entrepreneurs and innovators to ‘pitch’ their ideas to President Obama.

Yesterday saw the end of the ADA Initiative, “a feminist organization. We strive to serve the interests and needs of women in open technology and culture who are at the intersection of multiple forms of oppression, including disabled women, women of color, LBTQ women, and women from around the world.” (Ada Initiative, About Us)

How enormously ironic to see the closing of the one with the opening of the other, and both with such closely related missions. I can only hope that this first White House Demo Day proves to be one of many, and that the effort continues to embrace and support diversity as essential to American creativity and innovation.

White House Demo Day

The White House Demo Day had demonstrations to illustrate the diversity of people contributing to the innovation that helps strengthen the American economy. Most of the companies presenting had at least one woman founder or co-founder. Almost as many of the companies presenting had a founder that is a person of color or who shows ethnic or cultural diversity. The two companies represented by white men were (1) military, and (2) a winner of the XPRIZE. There were a few wonderful presenters from Michigan, including Ann-Marie Sastry of the University of Michigan Ann Arbor talking about her innovations in batteries and power storage. Products presented included new search engines based on cognitive models, medical innovations in cancer / HIV / aging / asthma, parenting tools, strategies for empowering patients, creative ways to repay student loans, several on converting ‘waste’ to profit, and much more. There was even Zoobean, who partner with libraries to recommend books and apps based on children’s preferences.

White House Demo Day

Part of what made this so wonderful (and why I wish I’d heard about it sooner) was the move to encourage parallel events across the country. I wish we’d done this here! Here are some tweets about the high points.

Read about the presenters here. Listen to the pitches here.


President Obama Hosts the First-Ever White House Demo Day https://www.youtube.com/watch?v=aKsxHS5vptM

White House Demo Day: https://www.whitehouse.gov/demo-day

Ada Initiative

“When the Ada Initiative was founded in 2011, the environment for women in open technology and culture was extremely hostile. Conference anti-harassment policies were rare outside of certain areas in fandom, and viewed as extremist attempts to muzzle free speech. Pornography in slides was a regular feature at many conferences in these areas, as were physical and sexual assault. Most open tech/culture communities didn’t have an understanding of basic feminist concepts like consent, tone policing, and intersectional oppression.” https://adainitiative.org/2015/08/announcing-the-shutdown-of-the-ada-initiative/

The Ada Initiative began by trying to change the world for women in STEM and tech. They stopped, but not without having made change, and not without leaving a permanent legacy. You’ll see tributes and comments below to testify to this, but you’ll also see links to some of the content they made open source and Creative Commons in order to help perpetuate their work, as well as work from some of their partners who carry on the good message and work. By the way, their open source toolkits are absolutely incredible and well worth downloading.

HOWTO design a code of conduct for your community https://adainitiative.org/2014/02/howto-design-a-code-of-conduct-for-your-community/
Code of conduct evaluations http://geekfeminism.wikia.com/wiki/Code_of_conduct_evaluations

Announcing the ADA Camp Toolkit: https://adainitiative.org/2015/07/add-a-little-bit-of-adacamp-to-your-event-announcing-the-adacamp-toolkit/

ADACamp Toolkit: https://adacamp.org/
– Inclusive event catering: https://adacamp.org/adacamp-toolkit/inclusive-event-catering/
– Providing conference childcare: https://adacamp.org/adacamp-toolkit/childcare/
– Quiet room: https://adacamp.org/adacamp-toolkit/quiet-room/
– Supporting d/Deaf and hard of hearing people at an unconference: https://adacamp.org/adacamp-toolkit/supporting-deaf-people/

Eszter Hargittai on Essential Tech Skills (#mlanet15)

Part 5 of a series of blogposts I wrote for the recent Annual Meeting of the Medical Library Association.


Eszter Hargittai #MLAnet15

Is the Internet good or bad? Safe or dangerous? The answer from Eszter Hargittai at the MLAnet15 Closing Plenary was, “It depends.” Even if I was tracking tech events at the conference, even if this topic wasn’t my own bread-&-butter, even though Natalie has already done an completely excellent blogpost about this session, I would (and do!) still want to blog on this.

“It depends,” is the correct answer to most debates about anything to do with the Internet. Eszter Hargittai has hard data, and strong methodologies behind it, to show that most of our stereotypes about Internet use, audiences, abilities, etc. are considerably more subtle, nuanced, granular, and shaded than commonly believed.

Eszter Hargittai #MLAnet15

In her research, there were a surprising or worrying number of students who were unable to distinguish phishing URLs from real URLs, to define fundamental Internet jargon terms, to determine accurate versus biased news, to leverage social sharing appropriately for employment, to find accurate and useful health information for common search topics, and more. She described impacts of lower socioeconomic status, gender differences that carry over to online environment, and importantly reframed the “digital divide” as “digital inequalities.” The example which most impressed me related to emergency contraception use. I suspect I’m not the only medical librarian who’s been approached by friends or neighbors with questions related to this topic. In Dr. Hargittai’s sample, a full third were unable to discover that it was possible to buy this over the counter at your neighborhood drug store. And that doesn’t even touch on the question of the other two thirds who do know how to find and purchase it, but do they know how to use it appropriately and safely in the context of their own health history?

Eszter Hargittai #MLAnet15

The Internet can empower as much as it can endanger. What makes the difference? Hargittai says skills.

Eszter Hargittai #MLAnet15

And the skills that are needed are ones that can largely be taught and learned. Librarians are people who generally (“It depends”, remember?) have both the needed skills and know how to teach them. This was a talk that the librarians who attended are still talking about, and I suspect that will continue in months to come.

You can explore the Storify, and its over 300 tweets to see what else people had to say about the talk.

GPII Gives me GOOSEBUMPS!

GPII Introductory Video

We all use technology every day. Well, people reading this blog, anyway. We use it, and everywhere we go, we either carry our own devices or spend time fiddling and fussing to make it work right. Or both. So, … have you heard about GPII?

GPII
Global Public Inclusive Infrastructure (GPII): http://www.gpii.net/

It gives me goosebumps, it really does. I’ve been hearing about this via Jane Vincent, author of “Making the Library Accessible to All” and a colleague here at the University of Michigan. Jane has been working on this project for a long time, before she came here. We are so very lucky to have her here and be informed literally at the ground level as this evolves. So what is it? What does “Global Public Inclusive Infrastructure” actually MEAN? It is the ultimate (for now) in portable technology personalization. Basically, how do you prefer to set up your computer? Now, code that into a little snippet, kind of like a credit card, and you take that with you wherever you go. Want to use a computer? Wave your magic card, and voilà! It’s set up just the way you like it.

I am, of course, oversimplifying, so here is a video introduction, and a video demo. Watch them both, and see if they don’t give you goosebumps, too!

Global Public Inclusive Infrastructure (GPII) w/captions and description:

Human Rights Museum Demo Instructions:

A big part of what I like about it is the community behind the idea and the process. Here’s a grateful nod to the following engaged and supporting organizations.

Cloud 4 All
FLOE Project
FLUID Project (which is worth an entire blogpost on their own!)
Prosperity 4 All
Raising the Floor
TRACE Center, University of Wisconsin

Personalized Medicine, Biosensors, Mobile Medical Apps, and More

At the Quantified Self Meetup, someone was praising the Rock Health slides. Of course, I had to go explore and see what was so great. These are my favorites.

About FDA’s Guidance for Mobile Medical Apps

FDA 101: A guide to the FDA for digital health entrepreneurs by @Rock_Health: http://www.slideshare.net/RockHealth/fda-101-a-guide-to-the-fda-for-digital-health-entrepreneurs

I especially took note of slide 10, where they describe things I would think of as an app, but which do not qualify as such for FDA regulation. This is an important distinction I hadn’t previously considered. Slide 12 takes it further by describing the categories of regulation as based on risk to patients, with good clear examples. Slie 21 on “pro tips” would have really benefitted companies like 23andMe (even though that isn’t actually a mobile medical app, the pro tips still apply, and in spades).

Biosensing Wearable Tech

The Future of Biosensing Wearables by @Rock_Health http://www.slideshare.net/RockHealth/the-future-of-biosensing-wearables-by-rockhealth

This one definitely gets into topics relevant to the quantified self movement and self-tracking. Slide six emphasizes the shift from the low hanging fruit (fitness, pulse, sleep) to the long tail — more targeted solutions for specific challenges (hydration, glucose, salinity, skin conductance, posture, oxygenation, heart rhythm, respiration, eyetracking, brain activity, etc.). That’s really quite interesting, and it gives examples of companies working in each space.

Slides 19-24 get into several of the areas our own local meetup defined as challenges to success for companies working in this space and for the future success of the entire area — it has to work, easily, and dependably. Slides 27-30 extrapolate these challenges into the transition into healthcare environments.

Personalized Medicine

The Future of Personalized Health Care: Predictive Analytics by @Rock_Health http://www.slideshare.net/RockHealth/the-future-of-personalized-health-care-predictive-analytics-press Video https://www.youtube.com/watch?v=UJak41hIDWc

SLIDES

VIDEO

It’s probably safe to say that most individuals working in the quantified self / self-tracking space eventually end up struggling with the issue of how to use their data to anticipate avoidable problems. This idea can be translated into the jargon phrase of “predictive analytics.” Slide 11 does a nice job of lining this up with how traditional healthcare is practiced, which is very useful. Slide 12 places this in the context of big data resources, databases, and tools, listing several of the main players. This context is essential for making personal data relevant beyond the drawn out process of n=1 studies. Slide 14 identifies the BIG problem of how companies working in this space largely focus on hospitals and health care providers, and seem to have entirely missed the idea that patients are deeply and actively engaged in this space. And, frankly, there are more of us than them (even if our pockets aren’t as deep). I love the phrase on slide 18, “Symptom calculators are the “recommendation engines” of health care.” Most of the rest of the deck identifies challenges and opportunities, which I hope any entrepreneurial types would examine closely. Do notice that there is a video with this one. You can hear the entire webinar as well as reviewing the slides.

Big Beautiful Questions (A HOTW post from #hcldr)

Guy with questions 8

The other blog for which I was writing the “Hashtag of the Week (HOTW)” posts has changed focus, so I am no longer doing them weekly, but I am still doing them when available time and something amazing both intersect. The something amazing part happens ALL THE TIME, and if that was the only factor, I could do these daily! But this time, the conversation was so relevant and useful that I would feel like I wasn’t doing my job if I didn’t share it.

Yesterday evening, the Healthcare Leadership group had a conversation about the role of questions and questioning in healthcare. The conversation was lead by Bernadette Keefe, MD, and was triggered by Warren Berger’s work in the area of “beautiful questions.” He wrote a book, but you can find a short intro to the core ideas in his New Year’s article, “Forget Resolutions.” To help people ask better questions, more answerable questions, questions that have a higher potential for leading to positive change in their life, Bernadette pointed out the tips from the “Right Question Institute“, and I pointed out the “Question Prompts List” strategy.

Right Question Institute Question Prompt Lists

The real value of the #hcldr conversation, however, came from the questions. The questions posed for the group, and the questions posted as answers. My favorite of the questions posed was, “What are we not asking?” Keep that in mind as you read the following selections from the questions given as answers to the prompts.


T1 In our sizable efforts to make healthcare more efficient, accurate and safe,as well as less costly, what are the questions weʼre not asking?


T2 As you experience healthcare delivery today – is questioning valued?


T3 What are questions you, personally, would like to ask of your healthcare provider, medical insurance company, or hospital?


T4 How could the value of questioning be incorporated into healthcare delivery in an efficient and effective way? Programs etc?


Closing Thoughts