Anthropic CEO claims AI models hallucinate less than humans
Amodei said all this in the midst of a larger point he was making: that AI hallucinations are not a limitation on Anthropic's path to AGI — AI systems with human-level intelligence or better.
"It really depends how you measure it, but I suspect that AI models probably hallucinate less than humans, but they hallucinate in more surprising ways," Amodei said, responding to TechCrunch's question.
Anthropic's CEO is one of the most bullish leaders in the industry on the prospect of AI models achieving AGI. In a widely circulated paper he wrote last year, Amodei said he believed AGI could arrive as soon as 2026. During Thursday's press briefing, the Anthropic CEO said he was seeing steady progress to that end, noting that "the water is rising everywhere."
"Everyone's always looking for these hard blocks on what [AI] can do," said Amodei. "They're nowhere to be seen. There's no such thing."
Other AI leaders believe hallucination presents a large obstacle to achieving AGI. Earlier this week, Google DeepMind CEO Demis Hassabis said today's AI models have too many "holes," and get too many obvious questions wrong. For example, earlier this month, a lawyer representing Anthropic was forced to apologize in court after they used Claude to create citations in a court filing, and the AI chatbot hallucinated and got names and titles wrong.
It's difficult to verify Amodei's claim, largely because most hallucination benchmarks pit AI models against each other; they don't compare models to humans. Certain techniques seem to be helping lower hallucination rates, such as giving AI models access to web search. Separately, some AI models, such as OpenAI's GPT-4.5, have notably lower hallucination rates on benchmarks compared to early generations of systems.
However, there's also evidence to suggest hallucinations are actually getting worse in advanced reasoning AI models. OpenAI's o3 and o4-mini models have higher hallucination rates than OpenAI's previous-gen reasoning models, and the company doesn't really understand why.
Later in the press briefing, Amodei pointed out that TV broadcasters, politicians, and humans in all types of professions make mistakes all the time. The fact that AI makes mistakes too is not a knock on its intelligence, according to Amodei. However, Anthropic's CEO acknowledged the confidence with which AI models present untrue things as facts might be a problem.
In fact, Anthropic has done a fair amount of research on the tendency for AI models to deceive humans, a problem that seemed especially prevalent in the company's recently launched Claude Opus 4. Apollo Research, a safety institute given early access to test the AI model, found that an early version of Claude Opus 4 exhibited a high tendency to scheme against humans and deceive them. Apollo went as far as to suggest Anthropic shouldn't have released that early model. Anthropic said it came up with some mitigations that appeared to address the issues Apollo raised.
Amodei's comments suggest that Anthropic may consider an AI model to be AGI, or equal to human-level intelligence, even if it still hallucinates. An AI that hallucinates may fall short of AGI by many people's definition, though.
This article originally appeared on TechCrunch at https://techcrunch.com/2025/05/22/anthropic-ceo-claims-ai-models-hallucinate-less-than-humans/
Try Our AI Features
Explore what Daily8 AI can do for you:
Comments
No comments yet...
Related Articles
Business Insider
27 minutes ago
- Business Insider
OpenAI employees share their 3 favorite tips for using ChatGPT
If you ever happen to see Nick Turley, the head of ChatGPT at OpenAI, muttering to himself on a weekday morning, it might be because he's talking to a chatbot. Turley said that ChatGPT's voice feature is his favorite tip for using the technology on a recent episode of the OpenAI podcast. "On my way to work, I'll use it to process my own thoughts. With some luck, and I think this works most days, I'll have the restructured list of to-dos by the time I actually get there," he said, adding that the voice feature isn't yet mainstream because there are a bunch of small "kinks" still. He said he finds it valuable to force himself to articulate his thoughts aloud, and wants to see the feature improve next year. Mark Chen, OpenAI's chief research officer, said on the podcast that he's a fan of the deep research feature, especially before an introduction. "When I go meet someone new, when I'm going to talk to someone about AI, I just preflight topics," Chen said. "I think the model can do a really good job of contextualizing who I am, who I'm about to meet, and what things we might find interesting." And podcast host Andrew Mayne, who was formerly OpenAI's science communicator and worked on ChatGPT, said he uses the technology when he's out at a restaurant. "I take a photograph of a menu and I'm like, 'Help me plan a meal or whatever, I'm trying to stick to a diet," Mayne said. Turley, however, cautioned against using the same trick for the wine list. "It keeps embarrassing me with hallucinated wine recommendations, and I go order it and they're like, 'Never heard of this one,'" he said. Corporate executives across companies are using AI in their daily lives, and OpenAI CEO Sam Altman is no different. Altman said on the "ReThinking" podcast in January that he uses it in "the boring ways," for things like processing emails and summarizing documents. When Altman spoke on the OpenAI podcast in June, he said that he uses ChatGPT "constantly" as a father. At the time, he said he was mainly using it to research developmental stages. "Clearly, people have been able to take care of babies without ChatGPT for a long time," Altman said. "I don't know how I would have done that."
Yahoo
an hour ago
- Yahoo
This is the Stock I'm Retiring On – It's Already Up 70%
According to estimates from S&P Global, just the five AI hyperscalers; Meta Platforms, Alphabet, Microsoft, Amazon, and Apple are projected to spend more than $1 trillion in capital investment on AI in the three years through 2027. Let that number sink in for a second…more than a trillion spent to develop artificial intelligence from just five companies. The AI revolution is NOT slowing down and stocks in the theme have the power to change your life! But while most investors chase new all-time highs in shares of Nvidia (NVDA), which would have to grow to more than the entire US economy to produce another 10X return, the smart money will be made in the next generation of AI benefactors. The companies in networking, software and final use cases – these are the stocks to buy now that will produce the next Nvidia-like returns. These are the stocks that will retire your job and I've found four I'm loading up on. First up, while everyone's been obsessed with Nvidia's chips, they've missed a critical truth: those chips are worthless without somewhere to live. That 'somewhere' is custom-built servers—and Super Micro Computer (SMCI) is the top supplier of those AI-optimized machines. Think of SMCI as the contractor building the actual brains of the AI revolution. They don't just sell off-the-shelf servers; they design tailor-made systems for speed, performance, and energy efficiency. And when companies are spending billions on data centers, every watt and millisecond matters. What gives SMCI its edge is time-to-market. It can go from idea to implementation faster than legacy giants like Dell or HPE. In a sector projected to attract over a trillion dollars in investment over the next two years, being first—and better—counts. The stock has already made some investors rich, but there's still plenty of upside in this picks-and-shovels powerhouse. Next up, ever wonder why Tony Stark (aka Iron Man) didn't just sell his AI assistant JARVIS out as a product and make trillions of dollars? SoundHound AI (SOUN) is doing just that, making voice AI actually useful in the real world. Its tech doesn't just turn voice into text and then figure out what you meant—it goes straight from speech to meaning, eliminating steps and reducing error. It's what Siri and Alexa should have been. SoundHound has spent over a decade perfecting this platform and is now embedding it into everyday experiences—from cars to drive-thrus and payment systems. Hyundai, Mercedes-Benz, Mastercard, and even White Castle are already rolling it out. These aren't pilot programs; this is real-world adoption, happening now. The market for voice AI is projected to $160 billion, potentially multiplying the company's $85 million revenue last year and turning this $3.9 billion company into a powerhouse. The biggest risk? That one of the tech giants—Amazon, Apple, or Google—decides it's easier to buy the company than compete before SoundHound AI has a chance to become the next Nvidia. I love talking stocks and that face-to-face community we're building on the YouTube channel. You can visit the Bow Tie Nation and check out all the 2025 stock picks on Let's Talk Money! Quantum Computing (QUBT) isn't in that AI theme…it's what could replace the theme and that's why I'm buying a few shares every month. When the market has forgotten about AI, when it becomes a part of our everyday life and investors look for the next big thing – it's going to be quantum computing and I want to be ready. Quantum computing isn't just faster—it's a complete reimagining of how information is processed. Traditional computers use binary logic—ones and zeros—while quantum computers leverage 'qubits,' which can exist in multiple states at once. The result? Problems that take supercomputers thousands of years could be solved in seconds. Quantum Computing Inc. is positioning itself as a software-first player in this revolution. While giants like Google and IBM are building massive, hardware-heavy quantum machines that require extreme cooling, QUBT is developing platform-agnostic software that works on today's systems but can seamlessly transition to quantum hardware when it's ready. Their Qatalyst platform abstracts away the complexity of quantum mechanics, which could be a game-changer for adoption. Symbotic Inc (SYM) is tackling one of the oldest and most broken systems in the economy, warehouse logistics and bringing it into the 21st century. Their solution combines AI, robotics, and software into a complete automation platform—something that transforms legacy warehouses into futuristic, self-operating fulfillment centers. The real genius here isn't just in the technology—it's in the business model. Symbotic doesn't sell a one-time system and walk away. Its tech is embedded deep into the DNA of a company's operations, making it nearly impossible to switch out without massive disruption for recurring revenue that keeps growing. They've already inked a 10-year deal with Walmart and expanded to Target and Albertsons. Now, through a $7.5 billion joint venture with SoftBank, they're offering automation-as-a-service to the entire industry. It's not just about U.S. dominance—the global logistics market is a $3.9 trillion opportunity, and Symbotic is just getting started. Shares are already up 70% this year but with runway to build the kind of portfolio that retires your job. Disclosure: This is the Stock I'm Retiring On is written by Joseph Hogue, CFA who is a former equity analyst and economist. Born and raised in Iowa, after serving in the Marine Corps, Joseph worked in corporate finance and real estate before starting a career in investment analysis. He has appeared on Bloomberg and CNBC and led a team of equity analysts for a venture capital research firm. He holds a master's degree in business and the Chartered Financial Analyst (CFA) designation. Positions in stocks mentioned: SOUN, SYM, SMCI Error in retrieving data Sign in to access your portfolio Error in retrieving data Error in retrieving data Error in retrieving data Error in retrieving data
Yahoo
an hour ago
- Yahoo
How AI Agents can transform banking operations: 3 principles for ‘endgame', not ‘game over'
AI Agents – a sophisticated type of software capable of planning, reasoning, and executing tasks independently – are fast becoming a serious consideration for banks looking to streamline operations and boost resilience. With organisations like the World Economic Forum (WEF) touting the transformational potential of Agentic AI, banking leaders must focus not only on the technology itself but getting it used effectively. Will AI Agents on balance displace banking jobs, or will they become integral to a new hybrid human-machine operating model? In other words, is it 'game over' for bankers or is this simply the 'endgame'? In the banking sector, where the cost of error is high and regulatory obligations are extensive, finding the answer hinges on more than just technological capability. Instead, it requires a clear understanding of how AI fits into existing systems, how it learns, and most crucially, how it understands the organisation it's deployed in. For AI Agents to be meaningfully integrated into core business operations, they need more than a generalised grasp of the world (what we might call a 'Public World Model'). Agents also require a 'Private World Model', a real-time, contextual understanding of the specific business environment they serve. This Private World Model is what enables AI to move beyond basic task automation and operate with the discretion, safety, and strategic alignment necessary for use in high-stakes settings like risk, compliance, or customer operations. Building it takes more than data. It takes a structured approach that brings business context into every layer of AI deployment. Banks seeking to move from early experimentation to strategic, at-scale adoption should follow these three key principles: For AI Agents to deliver value, they must be embedded into the operating model, not bolted on as isolated tools. That means defining their purpose, boundaries, and how they interact with human teams from the outset. In practice, this requires cross-functional alignment. That means bringing together risk, compliance, technology, and business operations to ensure governance is embedded and responsibilities are clearly allocated. It's about answering the operational questions before the technical ones. For example: What will the agent do? What decisions can it make? How will performance be measured? How will human oversight work? In highly regulated banking environments, this level of discipline is essential. Poorly integrated AI risks duplication, degradation of service quality, or worse, regulatory breaches and reputational harm. Successful AI programmes treat these issues as first-order design considerations, not afterthoughts. The temptation to adopt AI Agents quickly across the enterprise is understandable, but rarely effective. A more sustainable approach begins with well-defined use cases that offer a high return with manageable risk. One clear example is JPMorgan Chase's COiN (Contract Intelligence) platform, which uses AI to review commercial agreements. It reportedly cut error rates by 80% and freed up 360,000 hours of legal review time annually. This isn't a theoretical impact. It's measurable operational efficiency, delivered through structured implementation and ongoing oversight. Banks should look for similarly contained, repeatable tasks that are essential but burdensome. These create ideal environments for AI Agents to demonstrate value while allowing teams to build institutional knowledge and governance muscle before expanding into more complex areas. AI deployment is not a one-off exercise. As business needs change and regulatory frameworks evolve, AI Agents must adapt in parallel. That means embedding feedback loops and performance monitoring from day one. Unlike static software, AI systems learn from data, and that data changes. Ensuring AI Agents remain aligned with business strategy requires structured retraining, robust monitoring, and clearly defined escalation routes when things go wrong. Change management for the human workforce is equally important. As tasks evolve, new skills and new ways of working are needed. Supporting employees through this transition is critical to building trust in AI, ensuring adoption, and maintaining operational integrity. Retail banks must act now to embrace AI Agents before they become the industry standard, rather than a competitive edge. The prize is substantial for those who are first adopters: greater efficiency, faster decision-making, more consistent compliance, and more responsive customer operations. But the route to get there is not through a single piece of technology. It's through a deliberate strategy grounded in business context and operational clarity. By focusing on integration, strategic implementation, and continuous learning, banks can shift from seeing AI as a bolt-on and start treating it as a vital core capability. Rather than triggering 'game over' for bankers, AI's real potential lies in shaping a more agile, resilient and scalable workforce where humans and machines complement one another. That's an endgame worth striving for. David Bholat is Professional and Financial Services Director at Faculty "How AI Agents can transform banking operations: 3 principles for 'endgame', not 'game over'" was originally created and published by Retail Banker International, a GlobalData owned brand. The information on this site has been included in good faith for general informational purposes only. It is not intended to amount to advice on which you should rely, and we give no representation, warranty or guarantee, whether express or implied as to its accuracy or completeness. You must obtain professional or specialist advice before taking, or refraining from, any action on the basis of the content on our site.
