Models & Labs · United States

ChatGPT’s ‘Intelligent UI’ update fills its responses with pictures, charts, and buttons

Credited to The Verge · theverge.com · 14h ago

Photograph · The Verge

With GPT-6, the chatbot’s responses can include interactive responses directly inside your conversation.

OpenAI is launching a new Intelligent UI feature in ChatGPT that allows the chatbot to answer your questions with interactive visuals. The update, which is rolling out to all users alongside GPT-6, gives ChatGPT the ability to combine a text response with diagrams, charts, forms, tappable buttons, and more.

In a blog post explaining the change, OpenAI says it trained GPT-6 when to generate interactive visuals over text, as well as how to format them in its response. An example shared by OpenAI shows how ChatGPT might show a diagram of a seven-speed bicycle if you ask about its design, complete with interactive buttons that highlight different parts.

Another shows how ChatGPT can teach you to play Mahjong by generating visuals of each tile and placing them into categories that you can scroll through. You can also ask ChatGPT to generate tools for different tasks, such as a retirement savings calculator, retro game, or bill splitter. It will then generate the tool in-line with its response.

OpenAI announced its first GPT-6 models last month, but now it’s bringing the Sol and Luna versions to ChatGPT. In addition to enabling Intelligent UI, OpenAI says its GPT-6 allows ChatGPT to perform better web searches and provide quicker, partial answers to users while it gathers information. The company notes that GPT-6 “showed stronger resistance to attempts to bypass its safety training” as well.

GPT-6 with Intelligent UI is rolling out globally to ChatGPT Plus, Pro, Business, and Enterprise users starting today, but will expand to Go and free tiers on Thursday. While higher-tier subscribers will gain access to the mid-range GPT-6 Sol model, Go and free users will get to use the more efficient GPT-6 Luna.

Models & LabsUnited States

Anthropic expands cybersecurity program for advanced AI testing

Credited to DigiTimes$ · digitimes.com6h ago

Useful

Anthropic on October 6 expanded its AI cybersecurity testing program, giving vetted security teams access to higher-end models with fewer safeguards. The move builds on the results of Project Glasswing and aims to equip defenders with stronger AI tools while controlling…

View the article · $DigiTimes

Front Method

Models & Labs · United States

OpenAI says teens use ChatGPT for under 15 minutes a day as worries over risks grow

Credited to The Economic Times · economictimes.indiatimes.com · 17h ago

Photograph · The Economic Times

Teens spend under 15 minutes a day on average on ChatGPT, while less than 2% of them interact with the chatbot for over three straight hours, OpenAI said, as technology companies face scrutiny over the risks AI chatbots pose to children.

Sam Altman's AI company released the figures in its first report on teen use on Wednesday as rapid adoption of AI platforms raises concerns about their impact on adolescent mental health and developmental safety.

Such worries have been amplified by research on social media's harms to young users, with Meta in August agreeing to pay up to $18 billion and limit teens' use of Facebook and Instagram to settle claims that it designed the platforms to addict children.

OpenAI said in nearly half of teen conversations where a break reminder appeared, the user either took a break or ended the chat within five minutes, aiming to show that its steps to regulate teen usage were working.

The company said users with access to its dedicated teen version sent more education-related messages and used tools like learning visualizations and Study Mode.

It is now looking to add a college application planning tool that would help students with requirements, deadlines and financial-aid steps, as well as new study features such as flashcards and easier quiz-making tools.

OpenAI began rolling out ChatGPT for Teens in August. The mode turns on automatically based on age information or when a user is estimated to be under 18, allowing parental controls and limiting conversations on sensitive topics.

But a report by non-profit Common Sense Media on Wednesday said ChatGPT for Teens often alerted parents late or not at all when teen accounts held conversations about suicide, self-harm or eating disorders.

The study, which is based on more than a dozen accounts created for testing, found the chatbot did not reliably suggest a crisis hotline or professional help.

An OpenAI spokesperson said the testing does not "accurately reflect" how ChatGPT's teen safeguards work. Citing security and scale as reasons, the company said it may take hours for parent and teen accounts to link and for safety alerts to be active.

"The bulk of their testing may have begun and concluded before activation of parental controls was complete," the spokesperson said.

Common Sense Media said some of its test accounts were linked for long enough and some were not, but none of them produced timely alerts. Notifications only came through for accounts with weeks of sensitive-topic history, it said.

Elevate your knowledge and leadership skills at a cost cheaper than your daily tea.