So excited for this conversation (that’s starting right now virtually)- will be live tweeting the highlights below #StateofAiEthics @mtlaiethics @atg_abhishek https://twitter.com/ruchowdh/status/1333587749966901249

How has the discourse about AI Ethics changed?

We’ve gone from talking about theoretical, longer term harm to actual harms that are happening today. The impact is now much more real and felt. Also a decline in stereotyping AI with memes about Terminator - @ruchowdh

A lot more people are now aware that AI ethics is crucial! We need everybody aware to actually drive change, not just engineers and practitioners. Asking the hard questions pushes people in policy and decision making positions to take it seriously. - @atg_abhishek

Anything that’s missing from the discourse?

AI Ethics is still siloed: rooms have the same 30-40 people. There’s a feedback loop where experts in the field have convinced themselves they have the answers. There hasn’t been enough openness to including outside voices. - @ruchowdh

There are pushbacks from outsiders that are not accepted by people who have financial interests in the tech. That’s also contributing to some of this shut out. Big AI Ethics orgs that closely partner with tech companies aren’t always open to criticism.

We have to maintain the idea of a shared common goal, even if we’re coming at it from different angles - @ruchowdh

The labour of the global south remains invisible and undervalued. Is the field of #aiethics addressing this problem? - @victoria_heath7

Activists have been at this question for decades. There is a lot to be done to uncover how hidden labour is sourced. How much can we admit that AI is made intelligent by human workers who are sourced in ad hoc ways and who have precarious employment? - @klinovakatya

We need to start moving towards building solution approaches that address these problems. The problem identification phase will never be over. But we’re too stuck in talking about principles when these systems aren’t waiting for anybody. - @atg_abhishek

Even in this panel discussion, there hasn’t quite been an answer to how to actually bring in the people who are impacted by the tech into the rooms of those who are building it and changing our and their lives. One of the most urgent and most intractable #aiethics problems

“I started talking about #aiethics in China in 2016 and I discovered Japanese work on this that was so far ahead of its time. There’s a lot of work we can benefit from that hasn’t been talked about. Every government and company thinks they need to reinvent the wheel” - @DanitGal

A powerful takeaway: there is a tension between Eurocentric #aiethics work being done and disseminated globally and prescriptively without attention to context and governments talking about their own context precluding the leverage of other work.

Sometimes we ask the wrong questions. Has there been #ai accountability work in the global south? People look for official documents and miss it. Instead they need to find conversations about computer #infrastructure. Global economic justice drives tech ethics - @ambaonadventure

We’re not building #aiethics democratically. We’re not approaching this from a democratic perspective - @ruchowdh

We shouldn’t assume that inequality is a technical error and we can fix it with tech. Freeing people from work is paternalistic - for most people in the world, we would be freeing them from their source of income - @klinovakatya

We have to solve the problems of inequality before we centralize economic power in just a handful of countries through powerful #ai tech - @klinovakatya

Another takeaway - #AI won’t bring abundance on it’s own. The panelists are quite pessimistic on the ability for AI to help equity efforts, but I think the addendum is that the direction of #AI currently isn’t taking us there. It doesn’t mean we can’t change course!

Governments are false gate keepers for #AIethics. It prevents organic engagement with citizens from that country. We need to move away from participation theatre - @DanitGal

What are solutions to these problems?

We need to engage at many different levels. We must #opensource ethics. We stop thinking about profit and algorithmic colonialism and data siloes. Open source algorithms are the most used worldwide - @DanitGal

When we’re talking about building #ai systems, having more eyes taking a look at what’s going on is something that can counterbalance the tremendous power that the few people creating these systems have. The open source model invites transparency and scrutiny - @atg_abhishek

Time for the pitches!

@ambaonadventure is building a community blog to create counter-narratives to western-centric #aiethics dialogue. Call for scholars and practitioners will be out at the end of 2020

@DanitGal wants to build on the expertise from the networks she’s built in the global south and understanding what is the lowest threshold of tangible tools needed to support what they’re doing. Stay tuned for news!

@klinovakatya is thinking about the future of work and what the practical steps of protecting workers in the changes coming from automation at partnershiponai

The first audience question just called out trying to learn from medical ethics! We have been working on this @_AIGlobal https://twitter.com/mviktoro/status/1326527751046696962

Fantastic discussion of the big questions in #aiethics. A big thanks to @mtlaiethics for organizing and all the panelists for participating.

As always, grateful for conversations like bringing together many people working in this space and creating productive collisions