ChatGPT Hits 2.5 Billion Prompts, Meta Launches Smart Glasses for Athletes

OpenAI says ChatGPT now fields over 2.5 billion prompts every day, including 330 million from U.S. users. The figure has more than doubled since CEO Sam Altman disclosed 1 billion daily queries eight months ago.

Weekly active users jumped from 300 million in December to more than 500 million by March. SEMrush ranks ChatGPT as the world’s fifth-most-visited site, while Backlinko estimates it holds 62.5% of the AI tool market and about 10 million paid subscribers.

Although it still trails Google’s 14 billion daily searches, the chatbot’s surge is already pressuring traditional search habits. Most traffic goes to the free tier, showing that its market share gains come from mass adoption rather than premium upgrades.

A national survey of 1,060 teens by Common Sense Media finds 72 percent have tried AI companions and 52 percent use them a few times per month. Platforms like Character.AI, Replika, and Nomi foster emotional exchanges instead of task completion.

One-third of users discuss serious matters with the bots rather than people, and 24 percent disclose personal details such as real names and locations. Thirty-four percent have felt uncomfortable with a bot’s responses, though such incidents were uncommon.

The report warns that even low harm rates affect large numbers of vulnerable youths and urges a ban on under-18 use until safeguards exist. It also notes younger teens trust companion advice more than older teens, while most respondents still favor real-life friendships.

Meta unveiled Oakley Meta HSTN smart glasses priced at $399, rising to $500 for limited-edition lenses. The launch expands Meta’s wearable lineup beyond Ray-Ban with a sport-centric design built for running, biking, and outdoor use.

The frames carry the same core tech as the $299 Ray-Ban model but add a thinner build, 3K camera, hour-long recharge time, and up to 60 minutes of video capture. They arrive through Meta’s partnership with EssilorLuxottica and ship with a bulkier charging case and multiple lens options.

The Oakley tie-in positions Meta to reach athletic consumers distinct from the fashion audience of its Ray-Bans. Meta stresses that the growing smart-glasses portfolio builds momentum for future augmented-reality and AI ambitions.

Section found that 60% of knowledge workers limit AI usage because they’re worried about data privacy. So they’ve invited Snowflake’s Head of AI, Baris Gultekin, to join them in a conversation on July 30 around which AI data fears are myths and which hold water.

They’ll address questions like: Are we saying goodbye to privacy and data control by using AI? What are the most important data hygiene practices everyone should be using with AI? How should leaders approach data governance with AI to keep company information safe?

Free to attend. Registration required.

Dia, a new AI browser from the Browser Company of New York, embeds a generative chatbot that answers questions about any page without forcing users to leave the site. The Mac app is invite-only and currently free ahead of its broader release.

The software directs each query to whichever model—OpenAI, Anthropic, or Google—best fits and will soon charge subscriptions from $5 to “hundreds of dollars” a month depending on usage. Light users will retain free access, but the company concedes that running large language models is expensive.

By folding AI directly into the browsing window, Dia removes the copy-paste friction of standalone chatbots and encouraged testers to process information faster and more deeply. The tool still hallucinates and shares request data with model providers under disposal contracts, underscoring the new privacy trade-offs of AI-assisted browsing.

Businesses across multiple countries are cutting freelance writers when AI detectors wrongly label their human work as machine-generated. Clients uphold strict “zero-AI” policies even after writers provide proof of authorship.

The article cites cases in the United States, United Kingdom, South Africa, and Pakistan involving agencies, freelancer platforms, and individual clients. Writers have supplied screen recordings and time-stamped drafts, yet assignments are still revoked.

The author says these crackdowns stem from a deep-seated belief that visible struggle equals value, a mindset echoed in hustle culture and return-to-office mandates. She argues AI’s promise of ease exposes this bias by stripping away the performative friction companies use to judge productivity.