It might be good to reiterate (in part) why we’re all in here.
It might be good to reiterate (in part) why we’re all in here.
Any precedent here regardless of outcome will have significant (and dangerous) impact, as the status quo is already causing significant harm.
For example Meta/Facebook used to prioritize content that generates an angry face emoji (over that of a “like”) - - as it results in more engagement and revenue.
However the problem still exists. If you combat problematic content with a reply of your own (because you want to push back against hatred, misinformation, or disinformation) then they have even more incentiive to show similar content. And they justify it by saying “if you engaged with content, then you’ve clearly indicated that you WANT to engage with content like that”.
The financial incentives as they currently exist run counter to the public good
I asked AI the following question. I thought the response was worth sharing:
“What evidence is there for the merits of horseshoe theory?”
Let me analyze this carefully. The horseshoe theory suggests that the far-left and far-right of the political spectrum are more similar to each other than to the political center.
The main arguments made in favor of horseshoe theory include:
Both far-left and far-right movements have historically shown tendencies toward authoritarianism, with examples like Stalin’s USSR and Hitler’s Germany using similar tactics of state control, propaganda, and suppression of dissent.
Some similar rhetorical and organizational patterns appear in both far-left and far-right movements, such as: populist appeals, scapegoating of particular groups, calls for radical system change, and rejection of liberal democratic norms.
There are historical cases of individuals and groups moving between far-left and far-right positions, suggesting some underlying commonalities.
However, there are significant critiques of horseshoe theory:
It oversimplifies complex political ideologies and movements, forcing them into a rigid geometric metaphor that may not reflect reality.
Surface-level similarities in tactics or rhetoric don’t necessarily indicate deeper ideological similarities. The fundamental goals and values often remain drastically different.
The theory can be used to dismiss radical critiques of existing systems by equating all forms of political extremism, regardless of their specific content.
Political scientists generally view it as an oversimplified model that doesn’t adequately capture the multidimensional nature of political ideology.
Rather than a strict horseshoe pattern, many political scientists prefer more nuanced models that consider multiple axes of political belief and recognize that authoritarian tendencies can emerge across the political spectrum without necessarily indicating ideological similarity.
What are your thoughts on how political ideologies should be mapped or compared?