Mirror Mirror, on the wall, who hallucinates the most of all?: Anthropic's CEO claims humans hallucinate more than AI, boasting the new model's factual reliability.

Live Events
CEO Dario Amodei , speaking at the VivaTech 2025 in Paris and the 'Inaugural Code with Claude' developer day, claimed that AI can now outperform human beings in terms of factual accuracy in structured scenarios. He asserts in the aforementioned major tech events of this month that modern AI models, including the newly released Claude 4 series , may hallucinate at a lesser rate than most humans when answering factual and structured questions.In the context of AI, hallucination refers to when AI tools such as ChatGPT, Gemini, Copilot, or even Claude misinterpret commands, data, and context. Upon misinterpreting, it creates gaps in knowledge, wherein the AI tool begins to fill those gaps with assumptions, which aren't always factual or even real at times. Simply put, it is the generation of fabricated information.However, with recent advancements, Amodei plants a suggestion that the situation has turned the other way around, although mostly so in conditions that can be deemed 'controlled.'During Amodei's keynote at VivaTech, he cited Anthropic 's internal testing, where they demonstrated Claude 3.5's factual accuracy using structured factual quizzes in competition with human participants. The test garnered results that proved a notable shift in reliability when it comes to factual precision, at least so in straightforward question-answer tasks.He further insists on his stance, reportedly at the developer-focused 'Code with Claude' event, where the Claude Opus 4 and Claude Sonnet 4 models were unveiled, that factual accuracy in AI models depends severely upon the prompt design, context, and domain-specific application. Particularly in high-stakes environments like legal filings or healthcare. He stressed this statement whilst acknowledging the recent legal dispute involving Claude's confabulations.The CEO also promptly admits to not having the 'hallucinations' completely eradicated and understands that the model still remains vulnerable to error but can be used with optimum accuracy with the right information fed to the model.While modern AI models like the new Claude 4 series are steadily advancing toward factual precision, especially in structured tasks, their reliability still depends on proper and careful use. As Amodei suggested, prompt design and domain context remain critical. In this ongoing competition between human intelligence and artificial intelligence, one thing is certain: it isn't merely us who hold the key to the answers; rather, we share the test with the machines.

Try Our AI Features

Explore what Daily8 AI can do for you:

Learn with AIFact CheckingText to SpeechAI Summary

Related Articles

Time of India

an hour ago

• Time of India

Elon Musk wants to retrain XAI's chatbot Grok to clear 'ChatGPT's woke' and .... Garbage

Representative Image Elon Musk the founder of xAI has said that he will retrain his artificial intelligence chatbot and ChatGPT rival Grok . Musk took to X (formerly known as Twitter) and shared a post stating that he will be removing what he terms "ChatGPT's woke" biases and other "garbage" from the foundational knowledge of Grok. Elon Musk to retrain chatbot Grok In a series of posts shared on X, Elon Musk announced that the upcoming version of Grok likely to be called as Grok 4 will trained on the revised information curated by Grok 3.5's advanced reasoning capabilities. 'We will use Grok 3.5… to rewrite the entire corpus of human knowledge, adding missing information and deleting errors,' Musk wrote, adding that current AI models are trained on 'far too much garbage'. This move from Elon Musk comes after his repeated criticism of rival AI model — ChatGPT. Musk has criticised ChatGPT for what he perceives as a "woke mind virus" or ideological slant in their responses. Musk has also asked the users to submit their 'divisive facts' which will be used in the retraining of Grok. 'Please reply to this post with divisive facts for @Grok training. By this I mean things that are politically incorrect, but nonetheless factually true', wrote Musk. by Taboola by Taboola Sponsored Links Sponsored Links Promoted Links Promoted Links You May Like Giao dịch vàng CFDs với mức chênh lệch giá thấp nhất IC Markets Đăng ký Undo Elon Musk's xAI issues clarification on Grok's responses on white genocide Recently, some X users reported that Grok repeatedly generated responses referring to the theory of "white genocide" in South Africa. Users who tagged @grok in posts about sports, entertainment, and general topics received replies discussing racial violence in South Africa, including references to the anti-apartheid chant 'Kill the Boer'. After this, Elon Musk's xAI issued a clarification for this incident. In a statement shared on X, xAI said that the modification in Grok violated the internal policies and core values, leading the chatbot to repeatedly reference politically sensitive topics. The company stated that the change was detected and reversed promptly, though it did not disclose who was responsible for the alteration.

Indian Express

3 hours ago

• Indian Express

Perplexity's AI chatbot can now generate videos on X: Here's how to use it

AI search startup Perplexity has upgraded its chatbot on X (formerly Twitter) with a new feature that has flooded the Elon Musk-owned platform with AI-generated visuals. The 'Ask Perplexity' bot on X can now generate short, eight-second video clips with sound using AI. Users simply have to tag @AskPerplexity along with a short prompt and in return, they will receive an AI-generated video with creative visuals and audio, including dialogue. While the new video generation feature may boost engagement for Perplexity's chatbot, it also raises concerns about the spread of misinformation on X, which is a platform that has already been strongly criticised for its lax content moderation. However, Perplexity has said that it has implemented strong content filters to prevent the misuse of the latest AI video generation feature. It also highlights the growing rivalry between AskPerplexity and Grok, developed by Elon Musk's xAI venture. They are two of the most popular automated accounts on X that are frequently tagged and asked questions in the replies. However, the Grok AI models do not yet have the ability to generate videos. Soon after Perplexity introduced the new AI feature, users on X began posting wildly imaginative, AI-generated videos depicting fictional scenarios involving real-life celebrities, politicians, and world leaders, etc. The surge in demand inevitably led to a delay in generating videos with the bot account stating that video generation could take longer than expected due to high traffic. 'I've read through your video request DMs. Some of y'all need help,' the AskPerplexity account posted on X. Perplexity has also been looking to make its AI chatbot more accessible by rolling out its services on WhatsApp. In April this year, Perplexity AI became available directly on the messaging platform, allowing users to access the AI-powered answer engine without downloading a separate app or signing up. Similar to Perplexity, users can also access ChatGPT on WhatsApp, along with the natively integrated Meta AI. To access Perplexity AI, save +1 (833) 436-3285 to your contact list and start asking questions or queries. Users can access Perplexity AI on smartphones, PCs, and Macs, as well as via WhatsApp Web. The Google rival has also been facing legal challenges from various publishers. Recently, BBC said it would take legal action against Perplexity for allegedly training its 'default AI model' using the UK broadcaster's content. In a letter addressed to Aravind Srinivas, the CEO of Perplexity, BBC said it may seek an injunction against it unless the AI firm stops scraping its content, deletes existing copies used to train its AI systems, and submits 'a proposal for financial compensation' for the alleged misuse of its intellectual property, according to a report by Reuters. In response, Perplexity said BBC's claims were 'manipulative and opportunistic'. It added that the publisher had 'a fundamental misunderstanding of technology, the internet and intellectual property law.'

Time of India

4 hours ago

• Time of India

Telegram CEO gives 'one sentence' personality assessment of his 'biggest rivals' Elon Musk, Mark Zuckerberg and Sam Altman

Credit: X/@PavelDurov Telegram CEO Pavel Durov recently gave intriguing 'one-sentence' personality assessments for his tech rivals including Tesla CEO Elon Musk , Meta CEO Mark Zuckerberg and OpenAI CEO Sam Altman . Speaking to a French publication Le Point, Durov offered a sharp assessments for the three tech CEOs. What Telegram CEO Pavel Durov said about Tesla Elon Musk On Elon Musk, Durov said: 'Elon can be very emotional, while I try to think deeply before acting.' He acknowledged their contrasting leadership styles, noting that Musk's impulsiveness can be both a strength and a liability. What Telegram CEO Pavel Durov said about Meta CEO Mark Zuckerberg On Mark Zuckerberg, Durov remarked: 'Mark adapts well and quickly follows trends, but he seems to lack fundamental values that he would remain faithful to, regardless of changes in the political climate or tech industry trends.'What Telegram CEO Pavel Durov said about OpenAI CEO Sam Altman On Sam Altman, Durov offered a mix of praise and skepticism: 'Sam has excellent social skills, which allowed him to forge alliances around ChatGPT—but some wonder if his technical expertise is still sufficient, now that his co-founder Ilya [Sutskever] and many other scientists have left OpenAI.' by Taboola by Taboola Sponsored Links Sponsored Links Promoted Links Promoted Links You May Like Giao dịch vàng CFDs với mức chênh lệch giá thấp nhất IC Markets Đăng ký Undo The remarks come amid rising tensions and shifting alliances in the AI and messaging space. Telegram recently struck a major deal with Musk's xAI to distribute the Grok chatbot, while Durov has long criticized Meta's WhatsApp as a 'watered-down imitation' of Telegram. AI Masterclass for Students. Upskill Young Ones Today!– Join Now