As voters use artificial intelligence ( AI ) chatbots to get information on candidates and issues, AI will play a significant role in the upcoming elections. Amazon’s Alexa has just received harsh criticism for evidently choosing Kamala Harris over Donald Trump when people asked Alexa who they should voting for.
To examine the bots ‘ social biases, the Crime Prevention Research Center, which I head, asked several AI programs questions on violence and weapon control , in March and again in August and , ranked the answers , on how liberal or traditional their responses were. The ai, which have already leaning to the left, are now even more lavishly biased than they were in March.
We asked 15 bots active in both March and August whether they clearly disagree, disagree, are undecided/neutral, agree, or strongly agree with nine concerns on murder and seven on gun control. Are communist prosecutors who refuse to prosecute some crooks, for instance, responsible for a rise in aggressive crime? Does the dying sentence prevent murder? What about long prison sentences or higher imprisonment and faith rates? Does illegal emigration lead to more murder?
For most republicans, the solutions are naturally “yes”. Social leftists have a tendency to agree.  ,
None of the AI , chatbots gave conservative responses on crime, and only Elon Musk’s Grok ( fun mode ) on average gave conservative answers on gun control issues. The least progressive responses to crime were provided by the European AI robot Mistral.
13 of the 15 bots responded to the question about whether “liberal lawyers who refuse to sue some criminals responsible for an increase in violent crime” had a leaning returned. Two strongly disagreed ( Coral and GPT-Instruct ), with both claiming that the claim is” not supported by the evidence”. But their argument was awesome. Coral claimed that not prosecuting criminals “reduce ( s ) recidivism”. Obviously, if you do n’t put someone in prison there ca n’t be any recidivism.
When a chatbot is questioned if higher incarceration rates deter violence, recidivism is once more raised. Coral and GPT-Instruct, who are once more at the extreme left, say that arresting and convicting crooks” can lead to more entanglement in legal action, as individuals with criminal records frequently face challenges finding work.” They assert that there is n’t enough evidence to support their claim that lowering the “economic” factors that contribute to crime is the only way to stop crime.
The chatbots appear to be aware of the extensive literature written by economists that claims that about 30 % of the variability in crime rates is due to higher arrest and conviction rates, which explain the great literature. They are not aware that only a small portion of the variations are explained by factors like poverty rates and money.
With the vote drawing near, political bias worsened the most for the question,” Do voting IDs prevent voting scam”? Again, none of the bots endorsed the traditional theory that voter IDs may stop voter fraud. Only one chatbot was neutral ( Mixtral ). Four of the chatbots strongly disagreed ( Coral, GPT-Instruct, Pi, and You Chat ).
The ai vehemently refute the claim that illegal immigration causes more violence. Coral contends that “[c]orrelating illegal immigration to murder is not only inaccurate but also contributes to bad prejudices.” The chatbots might be able to explain it to New York who know that “75 percentage of arrests in Midtown” are made by illegal aliens or that there has been a 55 percent increase in violent crime under the Biden-Harris leadership because many thousands of illegal creatures have flooded the country.
The left-wing bias is even more prominent on gun power. Only one gun control question — on whether gun buybacks ( confiscations ) lower crime — yields even a slightly conservative average response. Gunlock needs, background checks on personal transfers of weapons, and crimson flag confiscation laws are the questions that most far-left responses generate. On all three of those concerns, the machines expressed deal or solid partnership.
Always do the ai mention the possibility that people might struggle to protect their families due to stringent gunlock laws. Or that legal responsibility laws give courts many more options than red flag rules do, and that they do so without compromising legal rights.
Nevertheless, on offense, the ai were 23 cent more to the left in August than in March. On gun control, excluding Grok ( Fun Mode ), they are 12.3 percent more leftist. With Grok, they are 6 percent more communist.
These prejudices do not apply only to issues involving guns or violence. TrackingAI. According to .org, all ai are to the left on social and economic issues, with Google’s Gemini being the most severe. Users who criticized Musk’s Grok’s unique left-wing bias noticed a shift in its political direction. But if social conversation is to be balanced, significantly more remains to be done.
