Reshaping Social Media: Stanford Study Supports Democratic Algorithms
Stanford HAI has studied how societal values can be embedded into social media algorithms.
This work hypothesises that even algorithms can be found to promote values in society, reducing partisan animosity by verifying the value approach based on political participation and mental health.
This study proposed a “societal objective function” to reduce partisan animosity in social media feeds. It also found that the values promoting democratic values in the social feeds would equally not decrease user engagements, just like those promoted by traditional engagement algorithms.
The premise of this line of inquiry for the job is to raise the retention of users on social media and maintain the democratic values therein.
This study showed several challenges, like resistance to content warnings or different effects related to the level of partisan strength of the user. The researchers express this intention to apply this research further in more comprehensive real-time experiments, narrowing down these societal objective functions in their application to various community norms.
A reflection is required on the possible impacts of the future scaling up of research applications in the imminent General Elections in several countries:
Besides social media algorithms’ ability to shape what appears on our timeline:
- Some news and other information could be classified as false and might affect targeted voter groups.
- Sensational content spreads faster, and this is what keeps people united. This may create confusion among voters,
- These algorithms create echo chambers reinforcing existing beliefs and deepening political divides.
The recent Stanford discovery could have significant implications for political elections. Here is how this approach would impact on the elections:
- Less partisan bias:
Prioritising societal values could at least lower some of the most partisan biases in the information landscape. That would represent a shift from algorithms currently in use, which amplify some political messages and hence reduce the possibilities of rigging voter opinions by pushing content that is either biased or skewed.
- Mitigation of misinformation and disinformation:
If the algorithms are reconfigured and designed to downrank or filter the content with misinformation, including fake news around political campaigns, they will be a critical step toward the integrity of the electoral process. This would address one of the prime risks prevailing in social media algorithms, where engagement often precedes accuracy.
-Reducing the creation of Echo Chambers:
Any new algorithms that are friendly to partisan hostility or promote content with the aim of social links and political participation could finally help dilute the echo chambers.
Regulatory compliance improvement:
The focus on ethical consideration and social benefits would be ideally positioned with emerging regulations that seek to reduce the negative implications of social media on elections through such algorithms. In the UK, US, and EU, various principles in the diverse regulations align with the goal of societal objective function. If explicated, it will shape new rules supporting democratic values when using algorithms.
I hope that applying the Stanford team’s approach to embedding democratic values into social media algorithms could help mitigate several risks associated with digital political campaigns.
Comments