Latest news with #Cluely
India Today
24-05-2025
- Business
- India Today
Anthropic will let job applicants use AI in interviews, while Claude plays moral watchdog
Anthropic, the AI startup behind the chatbot Claude, has officially walked back one of its most eyebrow-raising hiring policies. Until recently, if you fancied working at one of the world's leading AI companies, you weren't allowed to use AI in your application — particularly when writing the classic 'Why Anthropic?' essay. Yes, really. The company that's been championing AI adoption across industries had drawn the line at its own job candidates using it. But now, Anthropic's had a change of Friday, Mike Krieger, Anthropic chief product officer, confirmed to CNBC that the rule is being scrapped. 'We're having to evolve, even as the company at the forefront of a lot of this technology, around how we evaluate candidates,' he said. 'So our future interview loops will have much more of this ability to co-use AI.'Anthropic is changing its hiring approach"Are you able to use these tools effectively to solve problems?" Krieger said. He compared it to how teachers are rethinking assignments in the age of ChatGPT and Claude. The focus now is on how candidates interact with AI. For instance, what they ask it, what they do with the output, how they tweak it, and how aware they are of the tech's blind spots. This means that you can now bring AI along for the ride, but just be ready to explain how you played with Krieger made a solid point: if AI is going to be part of the job, especially in software engineering, then it makes sense to see how well candidates can use it, not ban it entirely. Another AI company, Cluely, also abides by the same rule. Know what it thinks, the policy shift, job postings on Anthropic's website were still clinging to the old rule as of Friday as reported by the Business Insider report. One listing read: 'While we encourage people to use AI systems during their role to help them work faster and more effectively, please do not use AI assistants during the application process.'Anthropic's hiring approach contradicts Claude 4 Opus AI moto, ethical AIWhile it seems pleasing to the eyes, it is in contrast to its latest Claude 4 Opus AI system. The model has been highlighted as a snitch. It's built to be super honest, even if it means ratting you out when you've tried something Bowman, an AI alignment researcher at Anthropic, recently shared on X (formerly Twitter) that the company's AI model, Claude, is programmed to take serious action if it detects highly unethical behaviour. 'If it thinks you're doing something egregiously immoral, for example, like faking data in a pharmaceutical trial,' Bowman wrote, 'it will use command-line tools to contact the press, contact regulators, try to lock you out of the relevant systems, or all of the above.'advertisementThis kind of vigilant behaviour reflects Anthropic's wider mission to build what it calls 'ethical' AI. According to the company's official system card, the latest version — Claude 4 Opus — has been trained to avoid contributing to any form of harm. It's reportedly grown so capable in internal tests that Anthropic has triggered 'AI Safety Level 3 Protections'. These safeguards are designed to block the model from responding to dangerous queries, such as how to build a biological weapon or engineer a lethal system has also been hardened to prevent exploitation by malicious actors, including terrorist groups. The whistleblowing feature appears to be a key part of this protective framework. While this type of behaviour isn't entirely new for Anthropic's models, Claude 4 Opus seems to take the initiative more readily than its predecessors, proactively flagging and responding to threats with a new level of assertiveness.
India Today
22-05-2025
- Business
- India Today
Creator of AI cheating tool says that technical job interviews for engineers are over, everyone will cheat
A month ago, a student at Columbia University made headlines, but for all the wrong reasons. Chungin "Roy" Lee was expelled from university and had his internships with Meta, Amazon, and TikTok revoked. The reason: he created an app, called Cluely, that helps engineers cheat in interviews. The story began when Lee posted a video on YouTube, showing off this app and how it works. While creating a viral app is a significant achievement, Lee landed on a disciplinary hearing. But this did not stop him from working on the app and making it even better. In his defence, he believes cheating with AI is the only fair way into the industry a recent interview with Business Insider, Lee stated, 'We say 'cheat on everything' because, ironically, we believe this is the only path towards a future that is truly fair." The statement gives birth to several ethical questions. One such question that is stuck in my mind is: if AI is the future and "cheating is the only way", is it even fair? If you visit the first thing that greets you is the large gray type that reads this. Lee says cheating will soon be standard practiceOnce known for creating software designed to assist job applicants in passing coding assessments using AI prompts, Lee has now expanded his ambitions. His company Cluely is positioned as an all-purpose tool that assists users during live conversations, from job interviews to first dates — even claiming to offer "cheating for literally everything."advertisement 'There's a very, very scary and quickly growing gap between people who use AI and people who moralise against it,' Lee said in an email to Business Insider. 'And that gap compounds: in productivity, education, opportunity, and wealth.'Lee believes that what's seen today as cheating will soon be standard practice. In another interview, he stated that once everyone begins relying on AI to navigate meetings, it will no longer be considered cheating — it will simply become the standard way people function and think moving predicts traditional interviews will become obsolete, replaced by AI-generated candidate profiles. These systems, he says, will analyse work history, skills, and compatibility to match candidates to jobs — leaving just a brief conversation to determine 'culture fit.''I already know all the work you've done, or at least the AI already knows the work you've done,' Lee told Business Insider. 'It knows how good it is. It knows what skills you're good at, and if there is a skill match, then I should just be able to match you directly to the job.'Lee reveals Cluely's hiring processCluely's own hiring process reflects this shift, with interviews reportedly replaced by informal chats. He said that since the company is not a believer of old-style interviews, it only aims to hold a conversation with the candidate. 'We check if you're a culture fit, we talk about past work you've done, and that's pretty much it," he the hiring process, Lee believes AI will fundamentally reshape the way people think, communicate and interact. In a new video, posted on EO YouTube channel, he said, "The entire way we're going to think will be changed."He added, 'Every single one of my thoughts is formulated by the information I have at this moment. But what happens when that information I have isn't just what's in my brain, but it's everything that humanity has ever collected and put online, ever?'He imagines a future where AI provides real-time summaries of people's lives, scraping digital footprints to give users condensed insights during interactions. 'What happens when AI literally helps me think in real time?' he asked. 'The entire way that humans will interact with each other, with the world, all of our thoughts will be changed.'Cluely, Lee says, is aimed at preparing people for this inevitable shift. 'The rate of societal progression will just expand and exponentiate significantly once everyone gets along to the fact that we're all using AI now,' he Lee, the divide between those who embrace AI and those who resist it will only grow. 'Mass adoption of AI is the only way to prevent the universe of the pro-AI class completely dominating the anti-AI class in every measurable and immeasurable outcome there is,' he told Business society accepts this vision or not, Lee is adamant: the AI revolution is already here, and it's time to keep up or be left behind.
Business Insider
21-05-2025
- Business
- Business Insider
Founder of AI tool for cheating in interviews predicts everyone will do it — and technical job interviews are on their way out
Creating a tool that allowed job candidates to cheat on their technical interviews kicked off a chain of events that would eventually see Chungin "Roy" Lee kicked out of Columbia University — but he believes that, soon enough, everyone will be using AI to get ahead. Lee has since branched out from an AI tool for coding interviews alone, founding "Cluely," which he's previously called " a cheating tool for literally everything," including live conversation. A promotional video for the app, for instance, depicts Lee using the app to "cheat" his way through a date. "There's a very, very scary and quickly growing gap between people who use AI and people who moralize against it," Lee told Business Insider in an email. "And that gap compounds: in productivity, education, opportunity, and wealth." "We say 'cheat on everything' because, ironically, we believe this is the only path towards a future that is truly fair," he added. Lee talked more about his views on how AI use will impact interviews in a recent interview with EO. "When every single person is using AI to cheat on meetings, then it's not that you're cheating anymore," he said. "This is just how humans will operate and think in the future." In the coming years, Lee expects interviews to be a lot more "holistic," and largely assess whether the candidate is a "culture fit," rather than focusing on a deep dive into their skills. That is, if the interview as a means of assessment endures at all, given that he expects AI to become powerful enough to build individual profiles for each candidate and feed that information back to the interviewer. "I already know all the work you've done, or at least the AI already knows the work you've done," he told EO. "It knows how good it is. It knows what skills you're good at, and if there is a skill match, then I should just be able to match you directly to the job, assuming that we get along after like a 30-minute conversation." It's a practice that's already commonplace at Cluely, Lee added, where he says interviews tend to be less formal. "I really don't know that there is a need for interviews in today's age, but right now what we use is really just a conversation," he said. "We check if you're a culture fit, we talk about past work you've done, and that's pretty much it." Lee expects AI to eventually alter more than just the job interview — he believes everyone will soon be using it as frequently and broadly as possible. "The entire way we're going to think will be changed," Lee told EO. "Every single one of my thoughts is formulated by the information I have at this moment. But what happens when that information I have isn't just what's in my brain, but it's everything that humanity has ever collected and put online, ever?" For instance, Lee posed — how different would an interaction between two people look if an AI could feed one a "condensed blurb" of information about the other, after it was finished scraping their entire digital footprint? "What happens when AI literally helps me think in real time?" Lee said. "The entire way that humans will interact with each other, with the world, all of our thoughts will be changed." With Cluely, Lee hopes to get people used to what he believes is an inevitable transformation. "The rate of societal progression will just expand and exponentiate significantly once everyone gets along to the fact that we're all using AI now," he said. "And that's what Cluely hopes to achieve, is to get everybody used to, 'We're all using AI now.'" For Lee, it's simple — either get on board or fall so far behind you can't ever catch up. "Mass adoption of AI is the only way to prevent the universe of the pro-AI class completely dominating the anti-AI class in every measurable and immeasurable outcome there is," he told BI.
Yahoo
03-05-2025
- Business
- Yahoo
Week in Review: Apple won't raise prices — yet
Welcome back to Week in Review! We've got lots of news for you today: Amazon earnings, Apple earnings, ChatGPT sycophancy, Alibaba's AI models, and much more. Let's get to it! Well? During Apple's earnings call yesterday, Tim Cook revealed that the company paid $900 million in tariffs last quarter. But despite this, the CEO didn't announce any price increases — yet. Big bucks: Defense tech startup Mach Industries, which was founded by 21-year-old Ethan Thornton, is about to close a fresh $100 million in financing, sources told TechCrunch. This new funding should bring the startup's total funding to about $185 million to date. More models: Alibaba announced this week that it's releasing a family of AI models, called Qwen3, that it claims can outperform OpenAI and Google. According to Alibaba, the Qwen3 models are 'hybrid' models, and they're not yet available for download. This is TechCrunch's Week in Review, where we recap the week's biggest news. Want this delivered as a newsletter to your inbox every Saturday? Sign up here. Shoot for the stars: Amazon's effort to create a space-based internet network began in earnest this week: The company fired off its first 27 internet satellites on Tuesday. The first half of the network needs to be deployed by mid-2026 to meet the FCC deadline. AI in your pocket: Meta is rolling out a stand-alone AI app, which it unveiled at the company's first LlamaCon event on Tuesday. The app allows users to access Meta AI in an app, similar to the ChatGPT app and other AI assistant apps. Blocked: A judge in India blocked the encrypted email provider Proton Mail across the country. This was in response to a legal complaint filed by New Delhi-based M Moser Design Associates, which alleged that its employees had received emails containing obscene and vulgar content sent via Proton Mail. Cheaters never win: Cluely went viral recently for its bold claim of helping people cheat on everything. But some startups are claiming they can catch Cluely's users. Alexa+: Amazon's new digital assistant aims to let users talk with it in a more natural style and eventually have agentic abilities that allow it to use third-party apps on a user's behalf. During Amazon's earnings call this week, CEO Andy Jassy says that 100,000 users now have Alexa+. AI where it matters: Airbnb seems to be taking a more measured approach with AI. It started rolling out an AI-powered customer service bot in the U.S. last month. CEO Brian Chesky said in February that the company would use AI for customer service before it started implementing it for other uses like travel planning or booking tickets. Epic ruling: Epic Games has notched a win in an ongoing legal dispute with Apple. Judge Yvonne Gonzalez Rogers said in a ruling that Apple was in "willful violation" of a 2021 injunction that prohibited the company from imposing anticompetitive pricing. As a result, Fortnite could be back on iPhones as soon as next week. LOL: Following the Epic Games ruling, Stripe shared documentation that shows iOS developers how to avoid the Apple commission. Check your settings: Meta emailed Ray-Ban Meta owners on Tuesday with a notice that AI features will now be enabled on the glasses by default. This means Meta's AI will analyze photos and videos taken with the glasses while certain AI features are switched on. Yikes! OpenAI said it was 'rolling back' the latest update to the default AI model powering ChatGPT, GPT-4o, after complaints about strange behavior, in particular it being overly agreeable and validating. According to OpenAI, the update, which was intended to make the model's default personality 'feel more intuitive and effective,' was informed too much by 'short-term feedback.' Now the company pledges to make changes that would prevent this from happening in the future. Oh the drama: We read through the 80-page Epic v. Apple decision to pull out the judge's juiciest, fieriest comments. Here's one to get you started: "Apple engaged in tactics to delay the proceedings. The Court later concluded that delay equaled profits." Sign in to access your portfolio
Yahoo
29-04-2025
- Business
- Yahoo
Startups launch products to catch people using AI cheating app Cluely
AI cheating startup Cluely went viral last week with bold claims that its hidden in-browser window is "undetectable" and can be used to "cheat on everything" from job interviews to exams. But some startups are claiming they can catch Cluely's users. And Cluely says it's ready to develop hardware products like smart glasses, or even brain chips, that bypass anti-cheating software altogether. San Francisco startup Validia launched a free product called 'Truely' last week in direct response to Cluely. The software triggers an alarm if it detects someone using Cluely, Validia says. Rhode Island-based startup Proctaroo also claims its platform can detect Cluely users. "When a Proctaroo session is active, we can see running applications and 'hidden' background processes – Cluely is no different," CEO Adrian Aamodt told TechCrunch, criticizing Cluely's business model as 'unethical.' Cluely's co-founder and CEO, Chungin 'Roy' Lee, called the anti-cheating tools promoted by these startups pointless, comparing them to years of failed cheating crackdowns in the video game industry. What's more, Lee says Cluely may go into hardware anyway, rendering anti-cheating software obsolete. 'Whether it's smart glasses, a transparent glass screen overlay, a recording necklace, or even a brain chip, we're not sure,' he said. Lee even says that expanding to hardware is 'quite trivial technologically," despite lots of high-profile AI hardware failures like Humane's AI Pin. Scrutiny of Cluely's business model does appear to have had some impact, though. Cluely has scrubbed references to cheating on exams and job interviews on its website and manifesto, a major original selling point. Now, Cluely only touts "cheating" on things like sales calls and meetings. Lee told TechCrunch that Cluely is 'redefining' its messaging to target the 'largest and most impactful markets.' 'Ultimately, we see a future where everyone uses AI to its utmost potential, and that means planting in large, specific markets, and expanding out from there,' he said. This article originally appeared on TechCrunch at
