Cointime

Download App
iOS & Android

OpenAI

ALL From OpenAI

Musk responds to OpenAI’s refusal to open source: The safest software is open source

Silicon Valley venture capitalist Mark Anderson forwarded a screenshot of an email from OpenAI co-founder Ilya Sutskever. Steven Sinofsky, former Microsoft executive and current board partner of venture capital firm Andreessen Horowitz, commented: "This is a classic argument against open source related to security. The idea that open source is insecure because 'bad guys' can see the source code to exploit weaknesses. Closed source can prevent this. This is completely wrong." Anderson responded: "The safest software in the world is open source. Most people pay attention and most bugs are fixed. QED (proof complete)." Musk responded: "Absolutely true."
AI

Singapore’s Temasek is in investment talks with OpenAI CEO Sam Altman

According to the Financial Times on March 5th, Singaporean state-owned investment company Temasek Holdings is in talks with Sam Altman, CEO of ChatGPT development company OpenAI, marking the first time a group with national backing has invested in OpenAI. Sources say that Temasek's senior executives have met with Altman several times in recent months to discuss preliminary but ongoing discussions, and no agreement has been reached on any investment scale. The talks come as Altman seeks funding to launch a semiconductor business aimed at reducing dependence on cutting-edge chips manufactured by Nvidia. Despite OpenAI's significant revenue growth since the launch of ChatGPT in November 2022, Altman has stated that OpenAI is still operating at a loss due to the enormous cost of building and training its models. Temasek's investment interest reflects its increasingly global footprint over the past decade, with the investment company currently holding a portfolio worth $287 billion globally, including some top-tier tech startups. OpenAI is currently supported by $1.3 billion from Microsoft (its largest investor), as well as several venture capital groups such as Thrive Capital, Sequoia Capital, Tiger Global Management, and Andreessen Horowitz.

OpenAI: categorically disagrees with the lawsuit filed by Musk

Based on an internal memo sent to company employees by OpenAI, the company "strongly disagrees" with Musk's lawsuit against the company. Musk previously filed a lawsuit with the San Francisco Superior Court, suing OpenAI and its CEO Sam Altman for violating the contract. Musk stated in the lawsuit that Altman and OpenAI violated an agreement made when the artificial intelligence research company was founded, to develop technology for the benefit of humanity rather than for profit.

THE MODELS ARE YOURS: THE PUBLIC'S LEVERAGE IN AI

On December 27, the New York Times filed a lawsuit, claiming that Microsoft and OpenAI infringed the Times’ copyrights by using its writings to train GPT-4 and other AI models. It follows a series of similar lawsuits from the Authors Guild, writers Michael Chabon and Sarah Silverman, and more.
THE MODELS ARE YOURS: THE PUBLIC'S LEVERAGE IN AI
AI

Italian regulator: OpenAI's ChatGPT violated privacy laws

Italy's privacy regulatory agency stated that OpenAI Inc.'s ChatGPT and technology used for collecting user data violate the country's privacy laws. The agency released the investigation results on Monday after launching an investigation in March last year. The results have been notified to OpenAI. The agency stated in the statement that it may constitute one or more violations of EU regulations. OpenAI has 30 days to respond. <br>

OpenAI CEO will meet with Samsung chip business CEO this week

According to a report from an unnamed industry insider cited by Korean Daily Economic News, Sam Altman, CEO of OpenAI, will meet with Samsung Electronics executives later this week in Seoul, including Co-CEO Kyung Kye-hyun. Altman may discuss various issues with Kyung Kye-hyun, who is responsible for the chip business, and other chip department executives, including artificial intelligence, supply of storage chips, artificial intelligence chip design capabilities, and investment in artificial intelligence chip production.
AI

OpenAI responds to the New York Times' infringement accusation: There is no legal basis and the training data comes from public channels

OpenAI has made a public response after being accused of infringement by The New York Times. It is reported that in order to train its powerful artificial intelligence (AI) language model, OpenAI will use large amounts of data collected from the internet, including Wikipedia articles, famous novels, social media posts, and all other content, and OpenAI has not obtained any permission. Microsoft is OpenAI's biggest supporter and has deployed OpenAI's artificial intelligence tools in several of its products. OpenAI responded to The New York Times' infringement allegations by saying that there is no legal basis for the allegations and that the training data comes from public channels.

Decentralized AI project Worldbrain completed a new round of financing, with Morgan Stanley participating in the investment

The decentralized artificial intelligence project WorldBrain has announced a new round of financing, with participation from Morgan Stanley, but the specific amount has not been disclosed. WorldBrain is an innovative project initiated by the Worldbrains Foundation, a subsidiary of OpenAI, which combines artificial intelligence, neuroscience, and blockchain technology to develop a comprehensive, multifunctional, decentralized artificial intelligence system. It is reported that this investment will promote WorldBrain's innovation in WorldModel, Web3 technology applications, and Depin facilities.
AI

OpenAI Negotiating Licensing Deals with Publishers Amidst Legal Turbulence

OpenAI is currently in talks with multiple publishers to secure licensing deals for their content in order to improve their AI models. The Chief of Intellectual Property and Content, Tom Rubin, has reported positive ongoing discussions with publishers. However, OpenAI is facing a lawsuit from The New York Times and Microsoft Corp. for allegedly using their articles without permission, which could result in significant damages and impact their content acquisition strategy. Despite this legal challenge, Rubin defended OpenAI's approach and emphasized that the content is only used for training models and not to reproduce or replace the original content.
AI

OpenAI pays media companies $1-5 million a year to train its large language models using news articles

Two media company executives who recently negotiated with OpenAI said that OpenAI pays between $1 million and $5 million annually to some media companies to use their news articles to train their large language models. Even for small publishers, this is a small amount, which may make it difficult for OpenAI to reach an agreement. Meanwhile, an executive said that Apple, which is trying to catch up with OpenAI and Google in the generative AI field, is also trying to reach agreements with publishers to use their content. Apple offers more funding, but also wants broader rights to use the content than OpenAI. (The Information)