◎ Burnt-faced Spoonbill
This Taipei mayor’s televised debate is the first large-scale application of AI in Taiwan for real-time fact checking and verbatim analysis of candidates’ speeches. The influence of AI can also be used as an important indicator to observe this election.
In the past, the election was about who could convince voters; now, the election also includes who can influence how AI answers voters.
Please continue reading
In the first televised debate of the Taipei mayoral election, Jiang Wanan and Shen Boyang, who were competing in the same debate, also broadcast the live broadcast on their personal YouTube channels. (File photo, taken from the "Shen Boyang Puma" YT channel) But there is a key premise. When AI conducts fact checks, it often gives priority to quoting public information from government agencies as the basis for factual judgments. In other words, if you ask AI about Taipei municipal issues, the information on the Taipei City Government website is crucial.
New battlefield for election polluting AI information sources
This is why I am so disgusted by the Chiang Wanan camp’s use of the Beijing Municipal Government as their campaign headquarters, forcing the bureau to continuously issue low-quality press releases such as “Slap Shen Boyang in the Face.” In addition to the low-level approach and violation of administrative neutrality, what is more serious is that this will cause long-term information misleading and pollute the source of AI information.
With the high degree of information fragmentation, the fragmentation of audiences, and the public's continued decline in trust in the media, the public's demand for relatively objective and credible information has become more intense and urgent in important occasions such as election debates.
Although some institutions or media have conducted relevant analysis in the past, such as the common debate lazy bag, it usually did not appear until the day after the debate, or even longer.
Debate live broadcast, the era of instant slap in the face is here
Today, AI is quickly filling this need, significantly shortening the time gap between information verification and analysis.
For example, during the live broadcast of the debate on October 31, there were many "instant slaps in the face" scenes. As soon as the debate ended, someone immediately compiled and analyzed the verbatim speeches of both parties, including the proportion of each party talking about municipal affairs and attacking their opponents, as well as the issues they focused on, recurring vocabulary, etc.
This kind of real-time and efficient information collection and analysis was also a rare phenomenon in past elections.
After the Taipei mayor's televised debate, AI was used on a large scale for real-time fact checking and verbatim analysis of candidates' speeches. The picture shows the Democratic Progressive Party's top ten lies exposing Chiang Wanan. (Taken from the Democratic Progressive Party Facebook) Although the current communication forms such as short videos and golden sentences are still the main way to strengthen the candidate's image label and incite emotions, with the assistance of AI, it will be easier for intermediate and independent voters who are interested in deeply understanding the candidate's policy propositions and language characteristics to obtain relatively complete analytical information as a reference for the final decision on voting intention.
I saw an AI verbatim analysis. When the two sides exchanged questions, Shen Boyang's questions were all focused on municipal affairs, while half of Jiang Wanan's questions were asking about Shen Boyang's past speeches. Looking at the overall debate, Jiang Wanan's attack rate on his opponents was three times higher than that of Shen Boyang.
At the same time as the debate, Jiang Wanan and Xu Qiaoxin suddenly came to life on Threads.
This kind of simultaneous operation of physical debates and online communities, combined with AI data analysis, it is obvious that Jiang Wanan’s camp set the goal of the debate to intensify the emotions of supporters and revive the momentum of the blue camp.
AI may gradually weaken the influence of KOLs
So far, it seems that it is indeed effective. But after the emotion subsides, it is still a question mark how many voters who originally held a wait-and-see attitude towards Jiang Wanan can be shaken.
But the problems with AI are also obvious. Just like the dilemma faced by Threads air combat, this AI information is likely to only further strengthen the echo chamber effect in the stratosphere.
In addition, the reason why opinion leaders in the media or the community have been able to exert an important influence in elections in the past is partly due to the relationship of authority and trust formed by the information gap.
But AI may gradually weaken the influence of these KOLs (key opinion leaders). In particular, those opinion leaders who have distinct political parties and have little professional reference value, but who always make political comments as "doctors", "scholars", etc., may end up just becoming combustion accelerators in the stratosphere.
There are many issues worthy of in-depth discussion in communication studies during this election campaign. Whether it is Threads as an emerging election air combat platform or AI's large-scale application in election campaigns, it will usher in an important empirical test after this election.
Jiang Wanan’s camp set the goal of the debate to intensify the emotions of his supporters and revive the momentum of the blue camp. (Taken from the Kuomintang Facebook) (The author is a former journalist)
