![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
abomvubuso posting in ![[community profile]](https://www.dreamwidth.org/img/silk/identity/community.png){kind=small}
talkpoliticsImages, videos, voice messages... in recent months, AI-generated content has caused a number of problems worldwide. For example, some photos that purported to show the arrest of Donald Trump, as well as some that were claimed to cover the war in the Middle East, have turned out to be created by artificial intelligence:
https://www.bbc.com/news/world-us-canada-65069316
The opportunities and risks that artificial intelligence creates have been the subject of a heated debate in political circles. And that's not surprising: in the coming year, there are many elections around the world to be held, including the decisive one for President of the United States and the vote for European Parliament. The EU wants to impose stricter rules on the use of AI, while some organisations are warning against over-regulating the market. In the meantime, an increasing portion of the general public now believe that AI is a threat to democracy.
We've seen it all this year: we've already witnessed false information spreading like fire during an election campaign, not without active help from AI. Before the elections in Slovakia, an audio generated by AI appeared and was distributed on FB and other social networks, purportedly featuring the voices of a major party leader and a journalist, discussing the manipulation of the upcoming election. It was not clear to users at first that the audio recording was a so-called "deepfake":
https://www.wired.co.uk/article/slovakia-election-deepfakes
It has become increasinfly evident that AI poses risks, especially for the formation of public opinion before elections. However, we are still at the very beginning of this process, so we can only imagine the direction in which events will develop in the longer run. Indeed, if put in the wrong hands, in addition to the targeted disinformation that can be spread by political and other groups, AI can also be used for entirely different purposes during elections.
In the social networks, for example, the use of countless bots pose a risk of influencing voters and distorting the election process. These are essentially fake accounts that are managed by AI systems and make the content more attractive to the algorithm so they can increase its exposition. This is often not obvious to users, so they get the impression that certain positions are more popular than others, at least online.
Nowadays personalised chatbots could also be used by politicians to interact with potential voters. In Miami, for example, Mayor Francis Suarez has created an AI chatbot to answer questions for him. Experts believe that such models could play a bigger role in the future:
https://apnews.com/article/florida-suarez-2024-campaign-artificial-intelligence-chatbot-6522b5802b8b33ee455c6c9eb367dd81
In Germany, most of the major parties are already using AI for political campaigning. In the last federal election for example, most parties used AI-based communication systems. CDU also launched an application that evaluated polling data.
Of course, the main purpose of such decisions is to increase the effectiveness of the election campaign. For example, with the help of AI, campaign workers can target those voters who are still on the fence.
In general, though, compared to other countries, European parties are quite cautious when it comes to using AI. The parties still really do not have reliable data about their own organisation and about their voters. Only when a party has well-processed data can it tailor its election campaign to it. The high level of data protection in most European countries prevents that from happening.
Thinking more long-term, some voices are urging the parties to agree on common guidelines regarding the use of AI at this point. These believe it would be good if the parties imposed some restrictions on each other. But so far this has not happened. Of course it's nice and all to be calling for more transparency from politicians about the purposes for which they use AI. But can it be done for real? I mean, in a meaningful way?
There are some practical arguments against it. In addition to opportunities for political parties, experts see AI as a chance for voters as well. They could use AI applications, for example, to get more information about candidates. Until then, however, the AI systems will need to be further optimised. The NGO "Algorithmwatch" is currently testing AI-powered chat models ahead of regional elections in Bavaria and Hesse in Germany. The chatbots sometimes gave completely wrong answers to questions about the various party candidates. According to "Algorithmwatch", this is a well-known problem with chatbots, as they simply calculate probabilities based on which they string together words - whether they are true or not. And I'm sure you know incorrect data can lead to incorrect answers:
https://algorithmwatch.org/en/bing-chat-election-2023/
While it's not yet possible to gauge exactly how AI is going to be used in the many upcoming elections next year, the topic will continue to be increasingly relevant, and just as urgent. Let's face it, all the transformation is out there, whether we like it or not. And technology is constantly developing. That's why we must constantly monitor these processes so as not to be surprised when AI completely transforms a given area, the election process and democracy most important of all.
https://www.bbc.com/news/world-us-canada-65069316
The opportunities and risks that artificial intelligence creates have been the subject of a heated debate in political circles. And that's not surprising: in the coming year, there are many elections around the world to be held, including the decisive one for President of the United States and the vote for European Parliament. The EU wants to impose stricter rules on the use of AI, while some organisations are warning against over-regulating the market. In the meantime, an increasing portion of the general public now believe that AI is a threat to democracy.
We've seen it all this year: we've already witnessed false information spreading like fire during an election campaign, not without active help from AI. Before the elections in Slovakia, an audio generated by AI appeared and was distributed on FB and other social networks, purportedly featuring the voices of a major party leader and a journalist, discussing the manipulation of the upcoming election. It was not clear to users at first that the audio recording was a so-called "deepfake":
https://www.wired.co.uk/article/slovakia-election-deepfakes
It has become increasinfly evident that AI poses risks, especially for the formation of public opinion before elections. However, we are still at the very beginning of this process, so we can only imagine the direction in which events will develop in the longer run. Indeed, if put in the wrong hands, in addition to the targeted disinformation that can be spread by political and other groups, AI can also be used for entirely different purposes during elections.
In the social networks, for example, the use of countless bots pose a risk of influencing voters and distorting the election process. These are essentially fake accounts that are managed by AI systems and make the content more attractive to the algorithm so they can increase its exposition. This is often not obvious to users, so they get the impression that certain positions are more popular than others, at least online.
Nowadays personalised chatbots could also be used by politicians to interact with potential voters. In Miami, for example, Mayor Francis Suarez has created an AI chatbot to answer questions for him. Experts believe that such models could play a bigger role in the future:
https://apnews.com/article/florida-suarez-2024-campaign-artificial-intelligence-chatbot-6522b5802b8b33ee455c6c9eb367dd81
In Germany, most of the major parties are already using AI for political campaigning. In the last federal election for example, most parties used AI-based communication systems. CDU also launched an application that evaluated polling data.
Of course, the main purpose of such decisions is to increase the effectiveness of the election campaign. For example, with the help of AI, campaign workers can target those voters who are still on the fence.
In general, though, compared to other countries, European parties are quite cautious when it comes to using AI. The parties still really do not have reliable data about their own organisation and about their voters. Only when a party has well-processed data can it tailor its election campaign to it. The high level of data protection in most European countries prevents that from happening.
Thinking more long-term, some voices are urging the parties to agree on common guidelines regarding the use of AI at this point. These believe it would be good if the parties imposed some restrictions on each other. But so far this has not happened. Of course it's nice and all to be calling for more transparency from politicians about the purposes for which they use AI. But can it be done for real? I mean, in a meaningful way?
There are some practical arguments against it. In addition to opportunities for political parties, experts see AI as a chance for voters as well. They could use AI applications, for example, to get more information about candidates. Until then, however, the AI systems will need to be further optimised. The NGO "Algorithmwatch" is currently testing AI-powered chat models ahead of regional elections in Bavaria and Hesse in Germany. The chatbots sometimes gave completely wrong answers to questions about the various party candidates. According to "Algorithmwatch", this is a well-known problem with chatbots, as they simply calculate probabilities based on which they string together words - whether they are true or not. And I'm sure you know incorrect data can lead to incorrect answers:
https://algorithmwatch.org/en/bing-chat-election-2023/
While it's not yet possible to gauge exactly how AI is going to be used in the many upcoming elections next year, the topic will continue to be increasingly relevant, and just as urgent. Let's face it, all the transformation is out there, whether we like it or not. And technology is constantly developing. That's why we must constantly monitor these processes so as not to be surprised when AI completely transforms a given area, the election process and democracy most important of all.
