After a quick photo op with the @DSP_SPE #scipol #NextGenCanSci #dreamteam, I'm attending the #CSPC2018 panel @sciencepolicy on the social implications of emerging technologies w/ @ideas_idees @SocMedDr @emmeslin @Scienceadvice

https://twitter.com/DSP_SPE/status/1060192634109489152

Discussion questions for 2-way Q&A at the emerging technologies panel #cspc2018 ⬇️

Directions: There are some questions that are easy to answer, others that are more important that are harder to answer (@emmeslin). The social implications of new tech, esp if algorithms can exponentially increase biases and misinformation, are the latter (@SocMedDr) #CSPC2018

#CSPC2018 @SocMedDr - most people are not literate in digital media, and so many engineers and entrepreneurs that design digital tools have little expertise in social implications. Better understanding from designers and users is necessary!

#CSPC2018 @emmeslin - the environment and framing of announcements, whether it's a new policy or #scicomm, are incredibly important and not always considered. We're good at naming and describing when it serves a political purpose, though.

Q on #SkillsGap and #DigitalLiteracy - historically most people don't know how to fix a car, how to blacksmith... why is the current skills gap different, why is it more important?
Dominic Martin @UQAM says there are inequalities in who controls the technology

. @SocMedDr adds that it's an issue of scale. The data industrial complex is huge, just look at the impact of social media on voting. When Zuckerberg goes to Congress and says they will build an algorithm to deal with #FakeNews, they need to be able to challenge and question this

Sandra Lapointe @McMasterU says in Q that we must distinguish the goals of (1) looking to definitively answer ethical questions and (2) take a decision on ethics for these technologies. @emmeslin says both are important and we do need to be comfortable with uncertainty #CSPC2018

#CSPC2018 @emmeslin goes on to define Closure = the sufficient threshold to call a question solved, for the time being.

Agreeing on what that threshold is, is a complex question to itself

So, next question, who is socially affected by new tech? Dominic Martin @UQAM says lower socio-economic classes and those playing lower functions in society will be more vulnerable to automation (though machine learning can take over high level tasks as well) #CSPC2018

Q from the back - ethics vary across cultures, and even within cultures. (the whole range is mentioned from Ayn Rand, to silicon valley thinking and extremists/terrorists). Q is, how do we ask ethical questions if ethics vary widely? Whoa, says panel at #CSPC2018 @sciencepolicy

#CSPC2018 @emmeslin says there's only so much one can do. There's risks and benefits to all approaches, and the history of how tech is governed has been when there is a mediatized event, e.g. One death during a gene therapy study set back the field by years.

Sarah Miller @MitacsCanada follows up on the morality question (esp for the most vulnerable), who is responsible for making decisions? @SocMedDr says this is a systemic question, to a systemic problem. Not everyone can make the choice to cut off from social media even over ethics

We need room for decision-making but this is just a piece, not the whole puzzle. Over-regulation is also an issue in several parts of the world - @SocMedDr #CSPC2018

It's also difficult to regulate tech that develops & changes so rapidly. Broad expertise needed (D Martin @UQAM)

Q from the front - w/ the increasingly rapid development of tech and its implication on training, schools and #PSE universities aren't able to adapt quickly enough. How do we deal with this? #CSPC2018

#CSPC2018 @emmeslin @Scienceadvice says it's challenging but many profs are adopting tech and introducing ethics from a grassroots level. One of his own grade 12 profs challenged their class about the science and ethics of in vitro fertilization ➡️ this shaped Eric's career.

#CSPC2018 @BelislePipon @HarvardU mentions the concept of ethics by design ➡️ how do we make sure the algorithms are good before they are deployed? Some designers say they do, but it's only to not break laws... How do we make sure they test for bias and social considerations?

Dominic Martin @UQAM says there's a lot to unpack, especially when bad algorithms build on each other in a cycle of garbage in, garbage out, garbage back in again #GIGO #CSPC2018

#CSPC2018 @emmeslin adds that much can be learned from going elsewhere, from other cultures. Economic development != ethical superiority or superior ways of knowing. @SocMedDr says training silos are not up to the task, interdisciplinary and transcultural exchange is necessary.

GREAT point from a student in the back. Students and people who don't have a seat at the higher table often need to mirror or follow the ethics of those higher up, to get that seat. It's tough for #nextgen to have an impact when their success depends on fitting the mold #CSPC2018

#CSPC2018 @emmeslin says the #NextGen are future leaders, and the old guard will step down soon, and it's probably overdue. Bringing different ways of knowing and diverse perspectives needs to happen.

Last Q, what can we do better to address social implications of emerging tech? D Martin @UQAM says we're trending towards openness in algorithms and how tech is built, but we do not understand all social implications yet. Lots more we need to know and consider #CSPC2018

#CSPC2018 @SocMedDr adds this is not just a policy problem, this affects everyone. @emmeslin ends by saying the impediment is not just communicating info about new tech to grow understanding, it's building capacity to critically assess them and engaging the public directly. Ask!