
The usage of generative AI successful online hunt is continuing to explode, moreover while galore group are dubious of nan technology's reliability and trustworthiness.
According to information first reported by Axios, ChatGPT now responds to astir 2.5 cardinal personification queries daily, pinch 330 cardinal of those (roughly 13%) originating successful nan US. That's astir 912.5 billion queries per year.
Also: How nan Trump management changed AI: A timeline
ChatGPT was besides nan most downloaded app successful nan world successful April; successful June, it clocked much App Store downloads than TikTok, Facebook, Instagram, and X combined.
All of those benchmarks person been reached since precocious 2022, erstwhile nan app was first publically launched. In little than 3 years, it has go an progressively cardinal portion of really galore group astir nan world prosecute pinch nan internet. OpenAI's chatbot remains a agelong measurement from surpassing Google -- nan hunt motor processes astir 5 trillion queries annually -- but ChatGPT could soon drawback up, particularly erstwhile OpenAI launches its long-awaited web browser.
(Disclosure: Ziff Davis, ZDNET's genitor company, revenge an April 2025 suit against OpenAI, alleging it infringed Ziff Davis copyrights successful training and operating its AI systems.)
Perhaps nan clearest grounds of nan seismic effect ChatGPT is having wrong nan online hunt manufacture has been Google's pivot, successful response, to its ain generative AI-powered features. AI Overviews has go nan astir conspicuous portion of users' relationship pinch Google's hunt engine.
Google is besides facing title from Perplexity, an online hunt start-up pinch generative AI astatine its core. Earlier this month, Perplexity launched its ain web browser, Comet, to compete pinch manufacture heavyweights for illustration Chrome and Apple's Safari. AI Mode, a characteristic Google debuted successful May that incorporates generative AI much profoundly passim nan hunt experience, operates rather likewise to Perplexity.
Trustworthiness and AI
Despite nan immense surge of generative AI wrong online search, galore group are still wary of nan responses they're receiving from AI.
According to a caller survey of much than 1,100 Americans, only 8.5% said they "always trust" nan accusation they person from Google's AI Overviews. Around 21%, successful contrast, said they person zero spot successful nan feature's expertise to aboveground meticulous information.
Also: 100 starring AI scientists representation way to much 'trustworthy, reliable, secure' AI
Trust issues notwithstanding, that aforesaid study besides recovered that much than 40% of respondents seldom aliases ne'er click connected accompanying web links provided successful AI Overviews to position nan root worldly informing nan AI model's response. The image that emerges is 1 successful which galore group are interacting pinch generative AI by default arsenic portion of their day-to-day Googling, while simultaneously placing small to nary credence successful nan answers they're receiving.
Another study published earlier this year, however, recovered that nan mostly of group spot ChatGPT much than quality experts successful astatine slightest 1 domain, but that spot diminishes erstwhile nan chatbot discusses delicate subjects for illustration ineligible aliases aesculapian advice.
Many different factors tin impact humans' perceptions of nan trustworthiness of AI chatbots, including nan reside of communication: More sycophantic AI-generated responses, for example, thin to be perceived as less trustworthy compared to responses that person a much neutral, little explicitly flattering tone.
While ample connection models are mostly trained to maximize facticity, they often veer unpredictably into mirage -- providing inaccurate accusation pinch nan veneer of truth.
Also: I said pinch an AI type of myself, acknowledgment to Hume's free instrumentality - really to effort it
Some starring AI developers -- for illustration OpenAI and Anthropic -- person launched initiatives aimed astatine interpreting and knowing nan intricate processes taking spot beneath nan hood of AI systems, some successful an effort to amended those systems' capacity and to boost personification assurance and trust.
Want much stories astir AI? Sign up for Innovation, our play newsletter.