Thank you Chairman Walden, Ranking Member Pallone, and the committee, for the opportunity to speak on behalf of Twitter to the American people. I look forward to our conversation about our commitment to impartiality, transparency, and accountability.
-
-
We build our policies and rules with a principle of impartiality: objective criteria, rather than on the basis of bias, prejudice, or preferring the benefit to one person over another for improper reasons. If we learn we failed to create impartial outcomes, we work hard to fix.
Show this thread -
In the spirit of accountability and transparency: recently we failed our intended impartiality. Our algorithms were unfairly filtering 600,000 accounts, including some members of Congress, from our search auto-complete and latest results. We fixed it. But how did it happen?
Show this thread -
Our technology was using a decision making criteria that considers the behavior of people following these accounts. We decided that wasn’t fair, and corrected. We‘ll always improve our technology and algorithms to drive healthier usage, and measure the impartiality of outcomes.
Show this thread -
Bias in algorithms is an important topic. Our responsibility is to understand, measure, and reduce accidental bias due to factors such as the quality of the data used to train our algorithms. This is an extremely complex challenge facing everyone applying artificial intelligence.
Show this thread -
For our part, machine learning teams at Twitter are experimenting with these techniques and developing roadmaps to ensure present and future machine learning models uphold a high standard when it comes to algorithmic fairness. It’s an important step towards ensuring impartiality.
Show this thread -
Looking at the data, we analyzed tweets sent by all members of the House and Senate, and found no statistically significant difference between the number of times a tweet by a Democrat is viewed versus a Republican, even after our ranking and filtering of tweets has been applied.
Show this thread -
Also, there’s a distinction we need to make clear. When people follow you, you’ve earned that audience. And we have a responsibility to make sure they can see your tweets. We do not have a responsibility, nor you a right, to amplify your tweets to audiences that don’t follow you.
Show this thread -
What our algorithms decide to show in shared spaces, like search results, is based on thousands of signals that constantly learn and evolve over time. Some of those signals are engagement, some are the number of abuse reports. We balance all of these to prevent gaming our system.
Show this thread -
We acknowledge the growing concern people have of the power held by companies like Twitter. We believe it’s dangerous to ask Twitter to regulate opinions or be the arbiter of truth. We’d rather be judged by the impartiality of outcomes, and criticized when we fail this principle.
Show this thread -
In closing, when I think of our work, I think of my mom and dad in St. Louis, a Democrat and a Republican. We had lots of frustrating and heated debates, but looking back, I appreciate I was able to hear and challenge different perspectives. And I appreciate I felt safe to do so.
Show this thread -
We believe Twitter helps people connect to something bigger than themselves, shows all the amazing things happening in the world, and all the things we need to acknowledge and address. We‘re constantly learning how to make it freer and healthier for all to participate. Thank you.
Show this thread
End of conversation
New conversation -
-
-
Behavior intending to silence another person? You mean like weaponizing your reporting system in order to mass report & get Conservatives suspended?
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
You know what’s a universal human right? Unchecked free expression.
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
1/ Wait. No. You can't just take the pin out of that grenade, pass it to your userbase, and run like mad- without expecting any consequences or for us to throw it back to you. This requires far more elaboration.
-
2/ "A default to free expression left unchecked can generate risks and dangers" Fine. What are those risks and dangers, and what are the risks and dangers of the alternative- silencing speech? Could it be that it contributes to echochamberification- division into camps of emnity?
-
3/"It’s important Twitter distinguishes between people’s opinions and behaviors" How is the distinction made? What does Twitter consider the parameters of each?
-
4/ "and disarms behavior" What are all forms of behavior that Twitter wishes to 'disarm', and how does it plan to do this?
-
5/'disarms behavior intending to silence another person' Problem 1: Twitter is making a determination of intent Problem 2: Twitter is making a determination of 'silencing'
-
6/ One could say that many posts criticizing person A's statement X is a raid-"behavior intending to silence" A, or anyone's expression of X. Another could say that what happened was simply expression countered with more expression. Where and how does Twitter draw the line?
-
7/ "disarms behavior intending to ... adversely interfere with their universal human rights." Problem 1: Twitter is making a determination of intent (how?) Problem 2: Twitter is making a determination of 'universal human rights'
-
8/ Rights are a very, very tricky business. For example, one could say that property rights, specifically to private ownership of the MOP, constitute a universal human right. Another could say that private ownership of even land constitutes theft, and a rights violation.
- 3 more replies
New conversation -
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.