| |  | Jan 22, 2026 | | | | | | Happy Thursday! Apple is developing an AI wearable pin. Anthropic lowers its gross margin projection to 40% from 50%. OpenAI CFO responds to Google's criticism of ChatGPT's move to introduce ads.
| | | | Apple is developing an AI-powered wearable pin that could be released as early as 2027, according to people with direct knowledge of the matter. The device would position Apple to compete more effectively with OpenAI, which is planning its own AI-powered devices, and Meta Platforms, which is already selling smart glasses. Apple's pin is a thin, flat, circular disc with an aluminum-and-glass shell and features two cameras—a standard lens and a wide-angle lens—on its front face, designed to capture photos and videos of the user's surroundings. It includes three microphones to pick up sounds in the area surrounding the person wearing it. It also has a speaker, a physical button along one of its edges and a magnetic inductive charging interface on its back, similar to the one used on the Apple Watch. Apple engineers are aiming to make the pin the same size as an AirTag, only slightly thicker. | | | | Anthropic last month projected it would generate a 40% gross profit margin from selling AI to businesses and application developers in 2025, 10 percentage points lower than its earlier optimistic expectations, The Information reported late Wednesday. The lower-than-expected gross profit margin resulted from the costs of running Anthropic models for paying customers, in a process known as inference, on servers from Google and Amazon. Those inference costs were 23% higher than the company had anticipated, the projections showed. Anthropic calculates gross margins by subtracting inference costs and other costs of selling its products. Its margin still improved from the year earlier. | | | | OpenAI CFO Sarah Friar said the company's custom inference chip is being "taped out"—the final step before manufacturing—and touted the progress of its more than $500 billion Stargate infrastructure project in a panel at the World Economic Forum Wednesday. In a conversation I moderated, Friar also pushed back against shots fired at OpenAI by Google Deep Mind CEO Demis Hassabis at Davos. In interviews with multiple media outlets, Hassabis said he was "surprised" that OpenAI was moving to introduce ads. In response to the criticism, Friar said: "I am trying to create as much optionality as possible." "Early is a weird word because in ad models you have to be at scale," she added. On Stargate, which was announced one year ago, she said that the project is "well over halfway built out" and OpenAI is already training models on servers at Oracle's Stargate campus. "It's gone better than we all dreamed," she said. | | | | Jeff Bezos' space company Blue Origin is planning to build an ultra-fast satellite internet service aimed at large companies, data centers and governments, the company said Wednesday. The service, called TeraWave, would put Blue Origin in competition with SpaceX's Starlink and Amazon Leo. Blue Origin plans to launch the first TeraWave satellites in the fourth quarter of 2027. Unlike Starlink or Leo, which are targeting both individual consumers and enterprise customers, TeraWave is aimed only at large customers who need especially fast internet speeds. Blue Origin said it plans to launch more than 5,000 satellites to support the service, which will provide internet speeds of up to 6 terabits per second. Currently, SpaceX's Starlink is by far the leader in satellite internet, with more than 9 million customers and billions of dollars in revenue. Amazon Leo has launched 180 satellites and began an "enterprise preview" in November, but has yet to make its service widely available. | | | | Alibaba Group is preparing to list its chipmaking unit T-Head, taking advantage of strong investor demand for companies developing AI processors to reduce China's reliance on Nvidia, Bloomberg reported. Alibaba will first restructure the unit as a business partly owned by employees, before exploring an initial public offering, the timing of which is uncertain, the report said, citing people familiar with the matter. Several Chinese chip designers, including Moore Threads, MetaX Integrated Circuits, Biren Technology and Iluvatar Corex, went public in stock exchanges in Hong Kong and Shanghai over the past month, as China seeks to indigenize the supply of AI chips. Baidu also is poised to list its chipmaking subsidiary Kunlunxin. | | | | Shopify has told merchants that use its commerce software that it will open up sales through AI chatbots like ChatGPT later this month, according to its site and communications with sellers. Merchants will pay OpenAI a 4% fee on sales made through ChatGPT's checkout feature, a Shopify spokesperson confirmed, on top of typical transaction fees Shopify charges such as payments processing. OpenAI, Google and Microsoft have all announced chatbot checkout features in recent months. Beginning Jan. 26, Shopify will begin making its merchants' products available for purchase through the checkout features in AI apps including ChatGPT, Google's AI Mode and Gemini and Microsoft Copilot, the company told merchants. As part of that rollout, most Shopify merchants' product details will be available to the AI apps through a data-sharing setup that Shopify announced late last year. Merchants' products will be available for purchase by default in AI checkout features that don't charge an additional fee. For now, sales made through Google and Microsoft's chatbot checkouts will not carry additional fees, the Shopify spokesperson said. For checkouts that charge a fee, such as ChatGPT, merchants will need to opt in to make their products available for purchase. Merchants can toggle AI platforms on or off individually to choose which channels to sell in, but their products will still appear in AI responses with links to their own websites. To keep products from appearing in AI responses, merchants must contact the AI companies or prevent their listings from being indexed by web crawlers. | | | | Recommendation app Yelp on Wednesday said it agreed to acquire AI agent startup Hatch for $300 million, or roughly 12 times the startup's annual recurring revenue. Hatch charges a monthly fee for AI-powered agents that can respond to customer inquiries and make appointments. Yelp, which has a $1.8 billion market capitalization, said it plans to operate Hatch as a standalone business as well as offer Hatch's services to plumbers, contractors and other professionals that list on Yelp. Automating such tasks is "an area that's been exploding rapidly pretty much since 2023 or so with the release of ChatGPT….that ignited the opportunity that Hatch was really a first mover on," said Yelp CEO Jeremy Stoppelman. Founded in 2018, New York-based Hatch employs about 150 people. It has raised about $14 million from Bessemer Venture Partners, Y Combinator and other investors. The deal is expected to close next month. Yelp said it is paying $270 million in cash and will spend $30 million for employee retention to be paid out over two to three years. Yelp said Hatch is generating $25 million annual recurring revenue, up 70% year over year. The purchase price as a multiple of forward revenue is less than some other recent AI startup acquisitions. Hatch's subscription revenue could diversify Yelp's business, which is largely based on selling advertising. Yelp in October launched its own AI agents which take calls on behalf of a service provider or restaurant. | | | | Amazon-owned primary care chain One Medical said it is launching an AI health assistant, a move that comes on the heels of similar healthcare chatbot announcements from Anthropic and OpenAI. One Medical members, who pay an annual fee for virtual and in person health care services, can use the assistant for medical advice, booking appointments and interpreting lab results. One Medical's chatbot uses models available on AWS service Bedrock. Amazon started testing the chatbot last year. Anthropic and OpenAI both said earlier this month that they were launching consumer healthcare services, and OpenAI also recently acquired Torch, a healthcare AI app that can collate medical data from various providers. | | | | Apple plans to bring more conversational capabilities to Siri, turning it more into a chatbot like ChatGPT, Bloomberg reported. The update, codenamed Campos, will take advantage of Google's Gemini AI model, part of a deal Apple announced with Google last week, the news outlet said. The Information also reported last week that Apple plans to bring more conversational capabilities to Siri using Google's technology. Bloomberg also reported that Apple is in talks with Google about using the company's cloud infrastructure, including its AI chips, tensor processing units, to launch the updated Siri. If those discussions turn into a deal, it would amount to a reversal of Apple's previous plans, which involved running its AI features only on Apple devices or in Apple's new private cloud system, called Private Cloud Compute. | | | | The Department of Justice on Wednesday named three people to a technical committee that will oversee the implementation of the remedy for Google's illegal search monopoly. The three are Tammy Savage, a former Microsoft general manager; Gerry Campbell, a longtime search executive who has worked at AOL and Alta Vista; and John Abowd, a professor emeritus of economics, statistics and data science at Cornell University, according to a filing from the DOJ. The technical committee is tasked with enforcing the remedy against Google, which includes banning exclusive distribution agreements for Google Search and forcing some data sharing with Google's rivals. The technical committee will consult on details of the remedy, such as how often Google has to share data and how search competitors can use search results syndicated from Google. The technical committee requirement mirrors the technical committee installed after Microsoft's antitrust loss, more than two decades ago. The three members named on Wednesday will select the remaining two members of the five-person committee. Savage, chosen by the Department of Justice, will chair the committee. Abowd, who was chosen by Google, has published research focusing on data security and privacy. During the remedies portion of the trial, Google repeatedly criticized plaintiffs' proposed remedies for posing data security concerns. | | | Popular articles
By Erin Woo and Stephanie Palazzolo
By Catherine Perloff and Sri Muppidi
| | | | | Opportunities Empower your teams to stay ahead of market trends with the most trusted tech journalism.
Learn more
Reach The Information's influential audience with your message. Connect with our team | | | | | |
0 comentários:
Postar um comentário