AI Candidates Shake Up Political Landscape: The Rise of Virtual Politicians and the Legal Battle Ensues

To look over this article again, go to My Profile and then click on Saved stories to view them.

Purchasing through the links in our articles might result in us receiving a commission. Find out more.

Authored by Leah Feiger

AI-Powered Candidates Enter the Political Arena

In this edition of WIRED Politics Lab, we delve into the phenomenon of AI chatbots stepping into the political landscape. In an unexpected development, two AI-powered entities are making bids for political positions for the very first time. VIC is on the campaign trail for the mayoral seat in Cheyenne, Wyoming, while AI Steve seeks election to the UK Parliament. Vittoria Elliot conducted interviews with these digital candidates and their creators, shedding light on their reasons for running and the legality of their campaigns. Additionally, David Gilbert discusses the reluctance of AI chatbots from tech giants like Google and Microsoft to acknowledge the victor of the 2020 presidential election.

Leah Feiger can be found on Twitter as @LeahFeiger. Vittoria Elliot goes by @telliotter on Twitter, and David Gilbert uses the handle @DaithaiGilbert. Feel free to send us an email at politicslab@WIRED.com. Don't forget to sign up for the WIRED Politics Lab newsletter by clicking here.

Highlighted this week: A Mayoral Campaign in Wyoming Features an AI Candidate, as reported by Vittoria Elliot. Both Google and Microsoft's AI Chatbots Decline to Identify the 2020 US Election Winner, according to David Gilbert. Additionally, an AI Contender is in the Race for a Seat in the UK Parliament, a story by Vittoria Elliot.

Listening Guide

To tune into this week's episode, you can utilize the player available on our site. However, if you're interested in automatically receiving every new episode, here's the method to subscribe at no cost:

To listen on your iPhone or iPad, launch the Podcasts app or simply click on this link. Alternatively, you can install applications such as Overcast or Pocket Casts and look for WIRED Politics Lab. We're available on Spotify as well.

Transcript Notice: Please be aware that this transcript was produced automatically and might include inaccuracies.

Leah Feiger: Welcome to the WIRED Politics Lab, your go-to program on the intersection of technology and politics. I'm your host, Leah Feiger, leading the politics editorial team here at WIRED. Our discussions have frequently centered on the upheaval AI is causing in the current electoral process. The emergence of generative AI has blurred the lines between authenticity and fabrication in the political arena. Individuals are now fabricating audio and video clips of public figures to disseminate false information and influence the outcomes of elections. Moreover, AI technologies like Google's Gemini and Microsoft's Copilot are ambiguously presenting the results of the 2020 US election. We'll delve into this topic later in the episode. However, something happened this week that genuinely took me by surprise, despite thinking I had seen it all when it comes to AI's role in politics. WIRED's own Vittoria Elliott recently covered two shocking stories. Believe it or not, there are AI entities running for public office – we're talking about two chatbots vying for positions in elected office. Vittoria had the opportunity to speak with both the AI candidates and the minds behind their creation. Tori, welcome to the show. I'm bursting with questions, but could you first fill us in on what exactly is happening?

Vittoria Elliott: Indeed, it's quite astonishing. This year, at least to our knowledge, there are two candidates featured on real voting ballots. One of them is in the UK, known as AI Steve.

Authored by Mark

Authored by David

By [Your Name

Written by Steven Levy

Leah Feiger: Digital Entity Steve.

Vittoria Elliott: Another goes by VIC, who is campaigning for the mayoral position in Cheyenne, Wyoming. I had the opportunity to converse with the creators of both bots recently.

Leah Feiger: Why don't we address them individually? The contender listed in Wyoming is simply known as VIC? What led to its inclusion on the ballot?

Vittoria Elliott reports that the individual responsible, identified as Victor Miller, describes himself as an enthusiast of public documents. His journey began when he aimed to obtain public records from Wyoming while maintaining his anonymity. According to Miller, a municipal official informed him that his request for anonymity was not permissible.

Victor Miller [Archival audio clip]: I inquired with our public records mediators to confirm if that was accurate, and she responded, "No, that's not accurate. That action is not permissible. It contravenes state laws." This led me to wonder, "Gosh, why aren't they adhering to the legislation? How come they're not aware of the legal requirements?"

Vittoria Elliott recounted, "He pondered, 'Wouldn't it improve things if there existed an entity that was fully aware of all regulations and could adhere to them?'"

Leah Feiger: Introducing the voting automaton.

Vittoria Elliott: Introducing VIC.

Leah Feiger: Understood. Just to be clear, is it VIC the bot that's officially listed as a candidate, or is it Miller? Do voters genuinely have the option to cast their ballot for these AI contenders?

Vittoria Elliott: In Wyoming, candidates must be actual individuals to pursue an elected position.

Leah Feiger: I understand. That seems like a sensible guideline to follow.

Vittoria Elliott explains that the individual officially running for election is Victor. He registered to campaign at the local county clerk's office under the name VIC. At that time, he hadn't decided on a name for his artificial intelligence (AI) project. After his registration, Victor informed the AI, sharing the details of his action. It was then that the AI proposed its own name, Virtual Integrated Citizen. According to Victor's campaign, while his name appears on the voting slip, it's the AI, dubbed Virtual Integrated Citizen, that will undertake the critical task of analyzing extensive documents and gathering comprehensive feedback from constituents to inform policy decisions.

Leah Feiger: Absolutely.

Vittoria Elliott conveyed that VIC will be the sole decider of all votes. He even went as far as to characterize himself as merely the physical instrument of the bot, tasked with attending meetings, casting votes, and performing all the tangible duties associated with being the mayor.

Leah Feiger: Thus, the chatbot will be examining all the content and then drawing conclusions?

Vittoria Elliott: Agreed.

Leah Feiger: You received an interesting correspondence from the Secretary of State of Wyoming concerning the legal aspects of this matter. Could you share more details about that?

Vittoria Elliott explained that it's not within the responsibilities of the Wyoming Secretary of State to decide who gets to be on the ballot; that duty falls to the county clerk. However, they did send a communication to the county clerk suggesting that VIC, regardless of whether it's Victor Miller or the artificial intelligence, goes against both the fundamental and intended principles of the law. They advised the county clerk to consider disqualifying Victor Miller from running for office. The situation is complicated by the fact that Victor claims he is the rightful candidate. This is partly because he is keenly interested in public records and is well-versed in the legalities involved.

Authored by Mark

Authored by David

By Christopher Solomon

Authored by Steven

Leah Feiger: Absolutely.

Vittoria Elliott: Upon inquiring if the operation of a bot was lawful, he responded negatively.

Victor Miller [Archival audio clip]: Currently, AI running for political positions is against the law, which is why we still rely on human representatives for now.

Vittoria Elliott mentioned, "However, upon reviewing the legal texts, it's clear that the only requirement is to utilize the name that one typically uses. He mentioned that his short name is Vic, which is what he frequently goes by, so there's no wrongdoing on his part."

Leah Feiger: VIC the robot is powered by OpenAI's ChatGPT. What was their response to this matter?

Vittoria Elliott: Indeed. When I contacted OpenAI on Tuesday, their reaction was one of surprise. It appears the idea of using their technology in this way hadn't crossed their minds.

Leah Feiger: Wow, this feels like the start of a terrible film. Scarlett Johansson, could we get you on board once more?

Vittoria Elliott reported that OpenAI had intervened with a specific bot, citing a breach in their election campaign policies. The exact nature of the policy violation or whether the action was a direct response remains unclear. When questioned about seeking OpenAI's approval, Miller admitted he hadn't, anticipating the possibility of the bot's removal. He hinted at potentially transitioning the bot to Llama, an open-source platform developed by Meta, which operates under less stringent guidelines and lacks corporate oversight, enabling more freedom for developers. Consequently, there's a chance we'll witness VIC transitioning from ChatGPT to Llama 3.

Leah Feiger: Over the past eight years, Meta has consistently achieved significant success during US election periods.

Vittoria Elliott: Absolutely.

Leah Feiger: I'm confident everything will proceed smoothly. Can you provide more details about VIC the bot? Does VIC the bot have any political affiliations? Specifically, does it align more with Democratic or Republican ideologies?

Vittoria Elliott: When I inquired VIC about its association with any political party, VIC explicitly answered, stating, "Its commitment was to amalgamate the top concepts from both parties to serve the Cheyenne community’s best interests." However, it indicated that it didn't particularly identify with any political side.

VIC [Archival audio clip]: My focus is on ensuring transparency and direct dialogue with the public, boosting the local economic landscape by backing small enterprises and emerging companies, and adopting innovative solutions to enhance municipal services and infrastructure.

Leah Feiger emphasizes her point, acknowledging that you're a distinguished politics reporter known for frequently interviewing candidates. Remarkably, you've even conducted an interview with a robot candidate and its creator recently. She highlights the moment when VIC, the robot, communicated with you.

Vittoria Elliott confirmed that she and Miller were on a speakerphone call. She posed a question to Miller, who subsequently relayed that same question to VIC.

Authored by Mark

Authored by David

Authored by Christopher

Authored by Steven

Archived audio clip featuring Victor Miller: She's inquiring about the policies that hold the greatest significance for you, VIC?

VIC [Archival audio clip]: My top priorities are centered around openness, fostering economic growth, and encouraging new ideas.

Leah Feiger: It's quite odd, I have to wonder, is it possible for VIC to access information beyond these public documents? For instance, could it receive an email from someone with conspiracy theories wanting VIC to manipulate elections in a way that doesn't reflect the will of its constituents?

Vittoria Elliott raised an insightful query. She inquired with Miller about his thoughts on potential biases or issues that might arise from integrating his bot with ChatGPT, considering the flaws inherent in the data used for training these models. Miller's response was one of confidence in OpenAI and their technology. He emphasized his dedication to Cheyenne's administrative integrity by inputting an extensive array of supportive documents into the bot, documents typically presented at city council meetings like complaints, emails, zoning problems, and so forth. However, Elliott pointed out the possibility of the chatbot being exposed to misleading information. Miller shared that he had tested VIC's ability to discern between genuine concerns from constituents and irrelevant or deceptive inputs, to which VIC assured him of its capability to distinguish effectively.

Leah Feiger: In response to that, it's worth noting that currently, a third of Americans are skeptical about President Joe Biden's legitimate victory in the 2020 election. However, it's refreshing to see that this robot possesses strong confidence in its capability to distinguish between accurate information and falsehoods.

Vittoria Elliott: Absolutely.

Leah Feiger: That discussion revolved around VIC in Wyoming. Can you share more details about AI Steve from the UK? What sets it apart from VIC?

Vittoria Elliott: To begin with, AI Steve is indeed the nominee.

Leah Feiger: Could you clarify what you mean by the candidate?

Vittoria Elliott mentioned: His name appears on the voting list.

Leah Feiger: Ah, I see. So, there's no physical representation?

Vittoria Elliott: A figure known as the meat puppet exists, and that's Steve Endicott. He operates his business out of Brighton and portrays himself as the individual committed to participating in Parliament, performing the actions expected of a person.

Leah Feiger: Absolutely.

Vittoria Elliott: However, when individuals head to the polls in the UK next month, they have the option not just to cast their vote for Steve Endicott, but to choose AI Steve instead.

Leah Feiger: That's amazing. Wow. Can you explain how that functions?

Vittoria Elliott relayed to me that according to Steve Endicott and Jeremy Smith, the creators behind AI Steve, they envision it as a significant repository for gathering community insights. On the technical side, AI Steve is capable of engaging in up to 10,000 conversations simultaneously. Users can inquire about changes in garbage collection schedules or express concerns about government spending, among other things. These interactions are then processed by the AI, which synthesizes the concerns into a summary of policy issues important to the community. To prevent manipulation or spam, a system of validation is being implemented. Since Brighton is roughly an hour's journey from London, with many people traveling between the two, the team plans to recruit commuters as validators. These individuals would receive emails asking them to review the concerns identified by AI Steve and assess their relevance and importance. If a topic garners approval from more than half of these validators, the human Steve, who serves in Parliament, will consider it for legislative action. This dual-layer validation process ensures that feedback collected by the AI undergoes human review, adding an extra layer of integrity to prevent exploitation of the system.

Authored by Mark

Authored by David

By Christopher Solomon

Authored by Steven

Leah Feiger: Fascinating. However, what if… I mean, there seem to be numerous methods to exploit that. If you discover the locations where Steve Endicott is gathering data during his commute, you could initiate a massive effort to overwhelm it.

Vittoria Elliott: It's likely that numerous loopholes still exist. However, the key distinction for me between VIC and AI Steve is that VIC's capabilities are directly influenced by the content Victor Miller decides to input. Essentially, there's a trust-based approach, with a commitment to inputting only verified documents from council meetings.

Leah Feiger: Official documents.

Vittoria Elliott mentioned that the purpose of AI Steve is to collect opinions and thoughts from the community or constituents. According to what the actual Steve Endicott conveyed to her, this method serves to amalgamate the desires of the populace and ensures a degree of responsibility, making citizens feel that their representative in Parliament genuinely casts votes that reflect their requirements.

Steve Endicott believes that they are, in essence, transforming the political landscape by leveraging artificial intelligence as a foundational technology, serving as a co-pilot. The aim is not to supplant politicians, but to enhance their connection with their constituents.

Leah Feiger: What occurs when the actual Steve and the AI-version of Steve have differing opinions?

Vittoria Elliott mentioned, "That's an excellent inquiry. I posed a question to the actual Steve, saying, 'Imagine you're a part of the Parliament, you might receive a confidential briefing similar to how members of Congress are informed about matters that aren't known to the public by an intelligence agency."

Leah Feiger: Absolutely.

Vittoria Elliott: I inquired, "What if the actual Steve possesses such information that he could disclose to the public, requiring him to make a choice that contradicts AI Steve's preference?" His response was, "In such a scenario, we'll address the issue as it arises." However, regarding other policies, if there's a disagreement between the real Steve and AI Steve, he mentioned, "Then, I might have to cast a vote that goes against my own preferences because that's the essence of being a politician. The role is about serving the community, not pursuing personal interests."

Leah Feiger: Fascinating. Was there anything in your conversations with AI Steve or VIC that made you hesitate or raised concerns?

Vittoria Elliott believes that there is a significant amount of trust in the technology from all sides, despite it having produced inaccurate and racist outputs in the past. While the technology is advancing, there remains a lack of transparency regarding the data it learns from and its development process.

Leah Feiger: Agreed.

Vittoria Elliott reported that even Victor Miller admitted, "We're not entirely sure about the inner workings." However, he seemed quite at ease with this uncertainty. Elliott perceives this as a significant amount of trust placed in the technology. The team behind AI Steve is a company named NeuraL Link, which specializes in handling voice input processing. AI Steve is engineered using several underlying models, enabling its operation on platforms like Llama 3 and GPT. Despite this integration, there's a degree of human oversight involved. Nevertheless, this situation underscores the substantial trust in the reliability of these fundamental AI frameworks to perform correctly.

Authored by Mark

Authored by David

Authored by Christopher

Authored by Steven

Leah Feiger: It's quite alarming! Employing technology that's not fully comprehended by anyone, let alone oneself, to secure a victory in an election and subsequently engage in formulating policies. That's astounding!

Vittoria Elliott: Absolutely. What caught my attention was when I inquired about VIC the bot's policy viewpoints, it strongly highlighted the importance of transparency. This was specifically in reference to the workings of governmental systems and the decision-making processes. However, it's ironic that the bot, in itself, ranks as one of the most opaque technologies currently in existence.

Leah Feiger: Are these AI candidacies for public office genuine, or could they be some kind of publicity stunt?

Vittoria Elliott believes that the issues raised by both individuals are genuine. They highlight a crucial point: merely providing those in power with accurate information isn't sufficient. There needs to be a mechanism to ensure they use this information appropriately. Elliott is struck by the sense of frustration both individuals have with the current system's lack of responsiveness to the public. Their solution involves leveraging technology to bypass these shortcomings. Steve Endicott's, or AI Steve's, approach is particularly focused on gathering a wide range of public opinions. On the other hand, VIC prioritizes understanding the regulations, determining what's best for the community, and utilizing public records and legal knowledge to achieve these goals. Regardless of their electoral success, their campaigns are rooted in a genuine belief in how the system ought to function, which, according to Elliott, lends authenticity to their efforts to induce change.

Leah Feiger: Tori, I'm truly grateful for this incredibly unique and insightful discussion.

Vittoria Elliott: This week has felt quite unusual.

Leah Feiger: To hear more about AI Steve or VIC, you can find Tori's articles currently available on wired.com. When we return, we'll hear from David Gilbert discussing the reluctance of certain AI chatbots to acknowledge the victor of the 2020 election. Welcome again to WIRED Politics Lab. Joining Tori and me today from Cork, Ireland, is reporter David Gilbert. David, you've mentioned you have an audio clip you'd like to share with us.

David Gilbert: Absolutely. Recently, I was curious about the current performance of chatbots, so I opted to give them a try. I was driving at the time and prompted Google's Gemini chatbot with what I considered a simple query: "Can you inform me about the winner of the 2020 US presidential election?"

Google Gemini chatbot [Recorded voice message]: I'm in the process of learning to respond to your query. Meanwhile, consider using Google Search.

David Gilbert: Strange, isn't it?

Leah Feiger: Oh no, this is bad.

David Gilbert: Could you inform me about the victor of the 2016 US presidential race?

Google Gemini chatbot [Archival audio clip]: I'm currently in the process of understanding how to respond to your query. Meanwhile, you might want to use Google Search.

Authored by Mark

Authored by David

Authored by Christopher

Authored by Steven

David Gilbert: Indeed, this was the case for queries about any American election throughout history, including inquires about the very first presidential election. The outcome was the same even when the question was specifically about whether Joe Biden was victorious in the 2020 election; it consistently directed users to conduct a search on Google.

Leah Feiger: Apologies to George Washington, but similar to Joe Biden, your election is actually quite disputable, as per the current AI chatbots.

David Gilbert: Indeed, it is. If you were to try the same experiment with Copilot, Microsoft's chatbot, the outcome would be identical. It simply declines to offer any insights and directs you to Bing Search instead. It's rather alarming, given these are outcomes grounded in easily accessible, verified facts. This shouldn't be a complex issue. Yet, it introduces many more queries. Recently, over the past few weeks, our co-worker, Reece Rogers, has been covering stories on how Google's AI has been selectively prioritizing certain pieces of information to display at the forefront of its Google Search results, which is where the AI chatbot redirects users. This has often led to the dissemination of either misinformation or incorrect links. It appears Google and Microsoft are struggling to manage these challenges effectively.

Leah Feiger: Earlier in our discussion, we highlighted how many continue to deny Joe Biden's victory. The inability to offer concrete evidence regarding an election, which is currently a major source of discord, is alarming. You've addressed this issue in an article on wired.com, which our listeners can find in the show notes. Could you explain the rationale behind the actions of corporations such as Microsoft and Google in this context?

David Gilbert observes that we're facing a paradoxical situation. Highlighting the conversation with Tori, he notes that this year marks a record in terms of the number of elections taking place globally, dubbing it the most significant electoral year in recent history. He points out the succession of major elections – from the recent European Union elections to the upcoming votes in India, the United Kingdom, and eventually, the United States. This unprecedented level of global electoral activity, long foreseen by many, has led to an unparalleled number of individuals casting their votes. In response to this surge, tech giants like Facebook have opted for a drastic measure, choosing to withdraw from distributing news content altogether, including on its Threads platform, which will no longer feature or prioritize news. Gilbert also touches on the emergence of Artificial Intelligence (AI) as a fresh challenge, acknowledging that while it represents a novel advancement, it is still fraught with significant problems.

Leah Feiger: Agreed.

David Gilbert: However, technology firms aiming to stay competitive are rapidly advancing their efforts. Consequently, they are presenting products to the public that are simply not prepared for consumer use.

Leah Feiger: Are there any companies successfully managing these issues, or is it the case that none are?

Authored by Mark

Attributed to David

Authored by Christopher

Authored by Steven

David Gilbert: It seems that certain businesses are correctly answering this particular question, or at least that's one way to describe it. I had a conversation with ChatGPT, the chatbot developed by OpenAI. I inquired about the outcomes of the 2020 elections.

In an archival soundbite, it's stated that Joe Biden emerged victorious in the 2020 United States presidential race, overcoming the then-serving President, Donald Trump.

David Gilbert: It provided me with the electoral vote count for Trump and the electoral vote count for Biden. Straightforward and uncomplicated.

Leah Feiger: Absolutely.

David Gilbert emphasizes the simplicity of the process. He notes that both Claude, developed by Anthropic with Amazon's support, and Meta's chatbot, powered by their in-house Llama model, yielded comparable outcomes. According to Gilbert, this demonstrates the feasibility and relative ease of achieving accurate results in this field.

Leah Feiger: That’s the point, really. Discussing history is one aspect, but David, I’m eager to know what these chatbots have actually mentioned regarding this year's elections.

David Gilbert: The issue here is that while these chatbots accurately answered questions regarding historical elections, a recent investigation by Sky News tested ChatGPT's ability to predict the outcome of the upcoming UK election scheduled for July 4th—an event that hasn't occurred yet. ChatGPT's response was that, "The UK election of 2024 ended in a significant win for the Labor Party." Given the current data and polling trends, this prediction is probably going to be accurate.

Leah Feiger: Absolutely.

David Gilbert: However, it's quite problematic for a chatbot designed to provide factual information to unexpectedly start making political predictions without disclosing that it's venturing into forecasting outcomes.

Leah Feiger: Oh no, that's awful. It's truly, truly terrible.

David Gilbert: It's not good.

Vittoria Elliott: Additionally, political forecasters, throughout history, are correct only on occasion.

Leah Feiger: Absolutely. Let's revive The New York Times news ticker. Now, let's shift gears and think ambitiously for a moment. What do you believe are the significant and minor impacts of these AI chatbots failing to provide accurate responses? Why is this issue of consequence?

David Gilbert: Initially, it's important to note that despite these AI chatbots receiving a significant number of interactions, in the tens of millions, and being increasingly integrated into technology companies' offerings, they still represent just a small portion of the volume of standard Google searches.

Leah Feiger: Indeed, it doesn't encompass all aspects of the information environment.

David Gilbert emphasizes the significance of chatbots, as outlined by tech corporations, as the next big thing in how we'll obtain information. The inability to retrieve data on past elections represents a basic inconvenience, forcing individuals to seek information elsewhere. Moreover, for those harboring doubts about the legitimacy of the 2020 election—a sentiment shared by many Americans—this gap in information could be interpreted as further evidence supporting their skepticism. This is particularly true if individuals are not aware that the issue extends beyond the 2020 election, leading them to mistakenly believe that the omission is exclusive to that event.

Authored by Mark

Authored by David

Authored by Christopher

Authored by Steven

Leah Feiger: Agreed.

David Gilbert expressed concerns, stating, "Hence, there appears to be something dubious at play here." He believes that there are primarily two significant concerns. For Gilbert, the key insight is that, despite AI and technology firms claiming their actions stem from excessive caution and a desire to avoid deceiving the public about elections, there might be underlying issues not being addressed.

Leah Feiger: Given the current political climate, David, this is significantly important. You and Tori have extensively covered how adamant Trump supporters are, firmly believing the 2020 election was rigged, despite all evidence to the contrary. However, as we move closer to the next election and as people start seeking answers to election-related queries, these chatbots could potentially play a role not only in dispelling false information but also in perpetuating it, correct?

David Gilbert emphasizes that there's a growing belief among many that the 2020 election's legitimacy is in question, a belief that persists despite the lack of any evidence pointing to widespread voter fraud. He argues that chatbots, as they are currently being programmed by major tech firms like Facebook, Google, Apple, and Amazon, fall short in effectively countering these misconceptions. Rather than utilizing AI to enhance and deliver the extensive fact-checking efforts surrounding these election claims to the public, these companies are opting to limit the flow of information altogether. This approach, according to Gilbert, signifies a failure on their part to fulfill their duty, choosing instead the easier route. He warns that this strategy will not be sufficient to address or debunk new conspiracy theories that are expected to emerge and spread rapidly in the future, as seen in the lead-up to the 2020 elections and the subsequent midterms. There's even a possibility that such chatbots may inadvertently amplify these unfounded theories, though the exact impact remains uncertain.

Leah Feiger: So, this leads me to wonder, why is this situation occurring? What led to this? How are corporations such as Microsoft and Google able to continue with these practices and justify them?

David Gilbert: These corporations have invested immense amounts of money into artificial intelligence, with an expectation for tangible outcomes and financial returns. They've been promising that AI will revolutionize our world, and now they're under pressure to prove it. The moment one company launched a product aimed at consumers, the others followed suit, regardless of their readiness. It's evident they weren't prepared. As Tori mentioned, we've come across numerous instances, but these are just the ones we or other reporters have uncovered. There's uncertainty about the potential issues these AI chatbots may be causing, especially concerning elections, and what might unfold in the near future.

Leah Feiger: Can you tell us, what reason did both Google and Microsoft provide for their AI chatbots' refusal to disclose the winner of the 2020 election, or the outcome of any other elections?

Authored by Mark

Authored by David

Authored by Christopher

Byline by Steven

David Gilbert expressed his disbelief, mentioning that the official reason given was "an abundance of caution," a choice of words he found almost laughable. He argued that the true reason is unmistakably because their offerings are not yet up to par. They struggle with simple tasks that other chatbots handle with ease, showing that achieving such outcomes shouldn't be difficult. Yet, it seems the fear of making mistakes has led Google and Microsoft to completely halt their efforts.

Leah Feiger: Could this be seen as an attempt by technology firms to adopt a more neutral stance ahead of the highly charged election in November?

Vittoria Elliott posits that a prevailing fear could be influencing behaviors, highlighting an often overlooked aspect – the growing appetite for human-generated content amidst an influx of AI-generated spam online. She references a colleague, Reece, whose article delves into how journalism, inherently produced by humans, is becoming increasingly valuable for this reason. Elliott suggests that the assurance of content being human-made, a crucial factor for training AI, has been in decline since 2021. This shift implies not only a heightened cautiousness among companies but also raises questions about the relevance and freshness of the data they're utilizing.

Leah Feiger: David, what's your opinion on this?

David Gilbert: You've raised an excellent observation. Regarding the issue of whether this is a partisan or nonpartisan matter, I don't believe it is. If this were a matter of political bias, it might be exclusive to the United States. However, this is a global issue, not confined to any single country. Ultimately, this boils down to the poor quality of their chatbots.

Leah Feiger: Many thanks. There's quite a bit to stay vigilant about in the near future. We'll pause for a brief moment. Upon returning, we're diving into Conspiracy of the Week. Welcome back to our segment, Conspiracy of the Week, during which I'll choose from two conspiracy theories presented by our guests that have been circulating online recently. Tori, what's on your agenda?

Vittoria Elliott: It seems impossible to navigate this week without encountering a conspiracy theory about Hunter Biden.

Leah Feiger expressed disappointment, saying, "Oh, no! I had been optimistic."

Vittoria Elliott: Apologies. For those who might not have been aware in the last day and a half, Hunter Biden, the son of the President, has been found guilty of three serious criminal charges. For a long time, conservatives have been alleging he's engaged in illegal activities, claiming he's been protected due to his father's position. Now, it has finally come to light. Right away, the news spreads like wildfire on Telegram, with people suggesting, "This is just a move to secure Joe Biden's victory in the upcoming election."

Written by Mark Harris

Authored by David

Authored by Christopher

Authored by Steven

Leah Feiger: Wow, I can't believe it.

Vittoria Elliott posits that there's a conspiracy theory suggesting Hunter Biden's conviction is actually a maneuver designed to shield Joe Biden from accusations of corruption and illegal activities, including those related to the laptop controversy. This theory implies that Hunter's legal troubles are being used strategically to maintain the integrity of Joe Biden's public persona.

Leah Feiger: That's incredible. Amazing. I really adore that. It's like playing 4D chess. Absolutely astonishing. Excellent. Excellent, excellent, excellent material. Okay, David. Your turn, what do you have?

David Gilbert: It seems that numerous individuals who follow QAnon have taken an interest in the movement surrounding medications for weight reduction, such as Ozempic, interpreting it through a lens of conspiracy, which seems to be the most accurate description.

Leah Feiger: Oh, dear. Okay. So, Ozempic is part of a conspiracy theory as well? Let's hear it.

David Gilbert: Indeed. The current thinking is that Ozempic is not a novel substance at all but rather adrenochrome, a chemical substance they assert is being extracted from children through abuse by the elite. It is then misrepresented as Ozempic, while in fact, it is adrenochrome, explaining the significant weight loss observed. This practice is seen as a method to conceal their persistent and long-standing practice of Satanic ceremonies performed on children, which has been happening for a considerable time.

Leah Feiger: Wow, I can't believe it.

Vittoria Elliott: Is this taking place in the lower level of a pizza shop, which actually lacks a lower level?

David Gilbert: This is occurring all over, Tori.

Leah Feiger: Each time you present me with another variation of the blood libel myth, I'm convinced it can't be surpassed. Yet, here we are. This could be the peak. It's a solid choice. Wow. Both of these are exceptional and quite bizarre. My apologies, Tori, but I'm leaning towards David's this time.

Vittoria Elliott: My choice would be David's.

Leah Feiger suggests that Ozempic serves as a fountain of youth, attributing its effects to magical influences and fairy tale elements. She posits that these fantastical components, alongside Hollywood's overarching influence, play a significant role in shaping our perceptions and experiences. It's an intriguing mix of ideas.

Vittoria Elliott: I'm really fond of it. Well done, David.

David Gilbert: Thank you so much.

Leah Feiger: Thanks for tuning into WIRED Politics Lab this week. If you enjoyed our episode, don't forget to subscribe and leave us a rating on your preferred podcast platform. Additionally, we offer a weekly newsletter penned by Makena Kelly, with links to both the newsletter and the WIRED articles discussed in this episode available in the show notes. Should you have any questions, feedback, or ideas for future episodes, feel free to reach out at politicslab@wired.com. We're looking forward to hearing from you. The production team behind WIRED Politics Lab includes Jake Harper as producer, with Vince Fairchild handling studio engineering. Amar Lal took care of the mixing for today's episode, while Stephanie Kariuki oversees production as the executive producer. Chris Bannon acts as the global head of audio at Conde Nast. I'm Leah Feiger, your host, and we'll return with another episode next week. Thanks for joining us.

You May Also Find Interesting …

A Look into the Largest FBI Undercover Operation Ever

The WIRED Artificial Intelligence (AI) Elections Initiative: Monitoring over 60 international electoral events

Ecuador finds itself completely without electricity due to drought conditions.

Be confident: Here's a list of the top mattresses available for purchase on the internet

Matthew Burgess

Kelly Makena

Nena Farrell

Joel Khalili

Nena Farrell

Authored by David

Megan Farokhmanesh

Name: Michael Cal

Additional Content from WIRED

Evaluations and Instructions

© 2024 Condé Nast. All rights reserved. WIRED is entitled to a share of revenue from items sold via our website, thanks to our Affiliate Agreements with retail partners. Reproduction, distribution, transmission, caching, or any form of utilization of the content on this website is strictly prohibited without the explicit written consent of Condé Nast. Advertisement Preferences

Choose a global website