Category: Top News

IBL News | New York

OpenAI avoided allowing its video model Sora—released on Monday and available to ChatGPT Pro and Plus paid users—to upload photos or footage of real people, as many users expected. The company said it would roll out that feature when it is safe.

Generative video is a powerful and controversial tool due to the growing number of fraud cases related to deepfakes worldwide.

“Early feedback from artists indicate that this is a powerful creative tool they value, but given the potential for abuse, we are not initially making it available to all users,” OpenAI wrote in a blog post.

“We know that this will be an ongoing challenge; we’re starting a little conservative,” pointed out Will Peebles, a member of OpenAI’s technical staff and a research lead on Sora, during a livestream presentation on Monday.

Among other measures from OpenAI to prevent misuse, Sora-generated videos contain metadata to show their provenance that abides by the C2PA technical standard.

Also, to fend off copyright complaints, OpenAI uses “prompt re-writing,” designed to trigger when a user attempts to generate a video in the style of a living artist.

Many artists have AI companies, including OpenAI, allegedly training on their works without permission.

OpenAI said Sora was trained using publicly available datasets, proprietary data accessed through its vendor partnerships, and custom sets developed in-house.

Early this year, ex-OpenAI CTO Mira Murati didn’t deny that Sora was trained on YouTube clips, violating the Google-owned streaming platform’s usage policy.

Sora can create multiple variations of video clips from a text prompt or image and edit existing videos via a Re-mix tool. A Storyboard interface lets users create video sequences; a Blend tool takes two videos and creates a new one that preserves elements of both; and Loop and Re-cut options allow creators to tweak further and edit their videos and scenes.

Sora is not included with ChatGPT Team, Enterprise, or Edu plans and is unavailable in the European Economic Area, the United Kingdom, or Switzerland.

Other tech companies working on AI, such as Meta and Microsoft, have also been forced to postpone product releases in the EU due to the continent’s complex data privacy regulations.

It is also not currently available to people under the age of 18.

Credits are required to generate videos with Sora. ChatGPT Plus and Pro plans provide 1,000 and 10,000 credits, respectively, which reset monthly.

480p videos generated with Sora cost 20 to 150 credits, 720p videos cost 30 to 540 credits, and 1080p videos cost 100 to 2,000 credits.

Credits reset monthly at midnight, don’t roll over, and expire at the end of each billing cycle.

By default, Sora videos are watermarked with a visual indicator in the lower-right-hand corner.

IBL News | New York

Anthropic added a Google Docs integration to its Claude.ai assistant. This feature allows users to access and reason about a document’s content from Google Docs within their chats and Projects.

This way, Claude can summarize long Google Docs and reference historical context from the files to inform decision-making or help with strategic planning.

This integration is available on the Claude Pro, Team, and Enterprise plans.

Another update allows Claude to match users’ communication and preferred way of writing.

Users can choose from these styles:

• Formal: clear and polished responses
• Concise: shorter and more direct responses
• Explanatory: educational responses for learning new concepts

Beyond these preset options, Claude can automatically generate custom styles and edit preferences as they evolve.

OpenAI’s ChatGPT and Google’s Gemini have similar features that allow users to tailor responses based on their writing style and tone. The Writing Tools feature in Apple Intelligence also provides presets with similar styles.

IBL News | New York

Perplexity.ai debuted last month a shopping feature that offers recommendations with search results and allows users to place an order without visiting a retailer’s website.

This feature, called Buy with Pro, allows paid customers in the U.S. to check for products from select merchants right on the Perplexity website or app.

With the move, Perplexity is taking on Google and Amazon to capture a portion of shopping search results.

The shopping recommendations aren’t sponsored, appear tailored to the search, and include products across Shopify.

To scale e-commerce operations, Perplexity launched a free Merchant Program for large retailers. It provides payment integrations and free API access.

With startups like Daydream, Deft, and Remark, AI-powered shopping searches have become a growing business area. Amazon debuted its AI assistant Rufus earlier this year.

These companies are betting that AI will help you find an item quickly, and you won’t need to spend much time looking for something.

IBL News | New York

NASA has teamed up with Microsoft to create an AI chatbot called Earth Copilot.

NASA’s Earth Copilot, which uses Azure OpenAI Service, has condensed NASA’s vast scientific geospatial information and answers questions about Earth. This data can help drive scientific discoveries, inform policy decisions, and support industries like agriculture, urban planning, and disaster response.

It lets users interact with NASA’s data repository through plain-language queries. They can ask questions such as “What was the impact of Hurricane Ian on Sanibel Island?” or “How did the COVID-19 pandemic affect air quality in the US?” AI will then retrieve relevant datasets, making the process seamless and intuitive.

The development of this AI prototype aligns with NASA’s Open Science initiative, which aims to make scientific research more transparent, inclusive, and collaborative.

At the moment, the NASA Earth Copilot is only available to NASA scientists and researchers to explore and test its capabilities.

After internal evaluations and testing, the NASA IMPACT team said they will explore its integration into the VEDA platform, which already offers access to some of the agency’s data.

IBL News | New York

IBM partnered with Meta to integrate open-source Llama into the WatsonX.ai platform and noticed strong enterprise adoption. One case pertains to Dun & Bradstreet customer operations.

IBM’s Watsonx.ai provides a next-generation enterprise studio for AI builders worldwide to train, validate, tune, and deploy AI models.

Other enterprise use cases mentioned are:

– 𝗕𝘂𝘀𝗶𝗻𝗲𝘀𝘀 𝗗𝗮𝘁𝗮: Dun & Bradstreet generates company summaries, revisiting millions of companies quarterly to enhance data pipelines.

– 𝗦𝗽𝗼𝗿𝘁𝘀: Sevilla FC’s Scout Advisor identifies and evaluates recruits, integrating scouting data and NLP.

– 𝗖𝘂𝘀𝘁𝗼𝗺𝗲𝗿 𝗜𝗻𝘁𝗲𝗿𝗮𝗰𝘁𝗶𝗼𝗻: AddAI’s Q&A chatbot, built, reduced unanswered queries by 50%.

– 𝗚𝗼𝘃𝗲𝗿𝗻𝗺𝗲𝗻𝘁 𝗘𝗳𝗳𝗶𝗰𝗶𝗲𝗻𝗰𝘆: Llama expedites document processing and legal research, reducing case prep time by 50%.

– 𝗛𝗥 𝗔𝘀𝘀𝗶𝘀𝘁𝗮𝗻𝗰𝗲: Telecom HR assistants achieved a 75% automation rate and multilingual support.

– 𝗖𝗹𝗮𝗶𝗺𝘀 𝗣𝗿𝗼𝗰𝗲𝘀𝘀𝗶𝗻𝗴: Financial services streamlined call summaries, saving an hour per case and enabling same-day reports.

– 𝗧𝗿𝗮𝗻𝘀𝗹𝗮𝘁𝗶𝗼𝗻: Local governments developed Q&A assistants with advanced translation capabilities.

– 𝗦𝗮𝗮𝗦: A SaaS vendor enhanced contract management with watsonx.ai, making AI tools widely accessible.

– 𝗧𝗮𝗹𝗲𝗻𝘁 𝗔𝗰𝗾𝘂𝗶𝘀𝗶𝘁𝗶𝗼𝗻: AI-driven HR automation cut costs by 90% for a talent firm.

– 𝗙𝗶𝗻𝗮𝗻𝗰𝗲: Real-time asset valuation and improved transaction monitoring achieved faster insights and analysis.

– 𝗜𝗻𝗳𝗼𝗿𝗺𝗮𝘁𝗶𝗼𝗻 𝗠𝗮𝗻𝗮𝗴𝗲𝗺𝗲𝗻𝘁: Automated document classification sped up archive digitization and data cataloging.

– 𝗠𝗲𝗱𝗶𝗮: A press agency automated article reviews, translations, and social media edits, saving editors 95% of their time.

– 𝗧𝗲𝗰𝗵 𝗮𝗻𝗱 𝗜𝗧: AI assistants reduced support ticket resolution times from hours to minutes, achieving 93% accuracy.

IBL News | New York

On Monday, Nvidia showed a new AI model that understands and generates sound as humans do.

Called Fugatto (Foundational Generative Audio Transformer Opus 1), this model generates or transforms any mix of music, voices, and sounds described with prompts using any combination of text and audio files.

However, Santa Clara, California-based Nvidia, the world’s largest supplier of chips and software for AI systems, said it is still debating whether and how to release it publicly.