As large language models become increasingly capable, there is growing concern that they may develop reasoning processes that are encoded or hidden from human oversight. To investigate whether current interpretability techniques can penetrate such encoded reasoning, we construct a controlled testbed by fine-tuning a reasoning model (DeepSeek-R1-Distill-Llama-70B) to perform chain-of-thought reasoning in ROT-13 encryption while maintaining intelligible English outputs. We evaluate mechanistic interpretability methods--in particular, logit lens analysis--on their ability to decode the model's hidden reasoning process using only internal activations. We show that logit lens can effectively translate encoded reasoning, with accuracy peaking in intermediate-to-late layers. Finally, we develop a fully unsupervised decoding pipeline that combines logit lens with automated paraphrasing, achieving substantial accuracy in reconstructing complete reasoning transcripts from internal model representations. These findings suggest that current mechanistic interpretability techniques may be more robust to simple forms of encoded reasoning than previously understood. Our work provides an initial framework for evaluating interpretability methods against models that reason in non-human-readable formats, contributing to the broader challenge of maintaining oversight over increasingly capable AI systems.

Wyoming man Victor Miller, who filed mayoral candidacy as AI bot "VIC," speaks out after OpenAI shuts down his account. Victor Miller, who had been running as an artificial intelligence-powered bot named "VIC" [Virtual Integrated Citizen] in Wyoming's capital city, conceded his bid to make technological political history on Wednesday. Miller received 327 votes, or about 3% of the total cast, in Cheyenne's nonpartisan mayoral primary on Tuesday night, according to Laramie County records. On Wednesday, Fox News Digital obtained a statement from Miller saying that he and VIC came up short in their bid to change the definition of political machine in the Cowboy State's capital city: "Today, I, Victor Miller, concede the Cheyenne mayoral race. As the first person to put artificial intelligence directly on the ballot, offering voters the novel choice of AI governance, our campaign has marked a historic moment in politics and technology," Miller said.

A mayoral candidate in Wyoming who proposed letting an artificial intelligence bot run the local government lost his race on Tuesday – by a lot. The candidate, Victor Miller, announced his run for mayor of Cheyenne earlier this year, and quickly made headlines after he decided to run with his customized ChatGPT bot, named Vic (Virtual Integrated Citizen), and declared his intention to govern in a hybrid format, in what experts say was a first for US political campaigns. Before the election on Tuesday, the AI bot "told" Your Wyoming Link that its role in the government would be to provide data-driven insights and innovative solutions for Cheyenne, while Miller would serve as the official mayor if chosen by voters and would ensure that "all actions are legally and practically executed". But ultimately, Miller and his bot only received 327 votes out of the 11,036 cast. On Tuesday evening, Miller conceded the race, and said in a statement the campaign was not about him as a candidate, but rather it was about "offering voters a groundbreaking option: the chance to elect an AI that would make 100% of the decisions in office".

Fox News chief political anchor Bret Baier has the latest on the pros and cons of the bombshell developments on'Special Report.' TOP OF THE CHARTS: The U.S. topped another study that looked at the fastest-developing artificial intelligence industries in the world, according to a new report. AI ON THE BALLOT: A librarian running as a nonpartisan candidate for mayor of Cheyenne, Wyoming, promises to allow an artificial intelligence bot created by OpenAI to govern the state's capital city. AI POWER PLAY: Google has its eye on the prize -- artificial intelligence -- and it's making a bold power play in the tech arena. The company's recent Made by Google event was more than just showcasing new technology.

In a bizarre turn of events, two AI chatbots are running for elected office for the first time--ever. VIC is campaigning for mayor in Cheyenne, Wyoming, and AI Steve is running for Parliament in the UK. Reporter Vittoria Elliot interviewed both of the bots and the people behind them. She explains their motivations, and if any of this is even legal. Meanwhile, reporter David Gilbert talks about how Google and Microsofts' AI chatbots are refusing to confirm who won the 2020 election.

Victor Miller is running for mayor of Cheyenne, Wyoming, with an unusual campaign promise: If elected, he will not be calling the shots--an AI bot will. VIC, the Virtual Integrated Citizen, is a ChatGPT-based chatbot that Miller created. And Miller says the bot has better ideas--and a better grasp of the law--than many people currently serving in government. "I realized that this entity is way smarter than me, and more importantly, way better than some of the outward-facing public servants I see," he says. According to Miller, VIC will make the decisions and Miller will be its "meat puppet," attending meetings, signing documents, and otherwise doing the corporeal job of running the city.

Large language models like ChatGPT have recently demonstrated impressive capabilities in natural language understanding and generation, enabling various applications including translation, essay writing, and chit-chatting. However, there is a concern that they can be misused for malicious purposes, such as fraud or denial-of-service attacks. Therefore, it is crucial to develop methods for detecting whether the party involved in a conversation is a bot or a human. In this paper, we propose a framework named FLAIR, Finding Large language model Authenticity via a single Inquiry and Response, to detect conversational bots in an online manner. Specifically, we target a single question scenario that can effectively differentiate human users from bots. The questions are divided into two categories: those that are easy for humans but difficult for bots (e.g., counting, substitution, positioning, noise filtering, and ASCII art), and those that are easy for bots but difficult for humans (e.g., memorization and computation). Our approach shows different strengths of these questions in their effectiveness, providing a new way for online service providers to protect themselves against nefarious activities and ensure that they are serving real users. We open-sourced our dataset on https://github.com/hongwang600/FLAIR and welcome contributions from the community to enrich such detection datasets.

As enterprises embrace digital transformation, many are expanding customer bases beyond the confines of their pre-pandemic demographics. For example, the cross-border ecommerce market is growing at double the rate of domestic ecommerce -- driven by consumers seeking brands unavailable in their home countries. According to Worldpay, 55% of online shoppers worldwide purchased from another country in 2020. But while digitization provides an opportunity for businesses to expand their target markets, many run up against the challenge of localizing their content for particular customer segments. It's true that the majority of customers prefer to buy products with information in their native language.

The fake news: A new weekly satire show from the creators of South Park is using deepfakes, or AI-synthesized media, to poke fun at some of the most important topics of our time. Called Sassy Justice, the show is hosted by the character Fred Sassy, a reporter for the local news station in Cheyenne, Wyoming, who sports a deepfaked face of president Trump, though a completely different voice, hair style, and persona. Meta commentary: The first episode, released on YouTube on October 26, took on the topic of deepfakes themselves, with Fred Sassy warning his faithful viewers that they shouldn't believe everything they see. The satirical twist is that all the footage shown as real is, of course, deepfaked, while all the footage labeled fake is either real or played by puppets. The episode features a wide range of highly convincing deepfakes representing people including former vice president Al Gore, Facebook CEO Mark Zuckerberg, and president Trump's son-in-law Jared Kushner, whose face is deepfaked onto a child.

The door to education is communication, some say. And the big technology companies are opening that door more widely than ever before, not only as their products become more accessible to people with specialized needs, but also as educators find more ways to use those features in the classroom and beyond. The biggest technology companies -- think Apple, Google and Microsoft -- include language accessibility tools in their vast array of products, and those features are available to pretty much any user. And the companies have been actively improving those tools in recent years. Each of the tech giants has a group that promotes accessibility, said Luis Perez, technical assistance specialist at the National Center on Accessible Educational Materials, who himself has a visual disability.