Put simply Joe Biden and Vice President Kamala Harris met this week with leaders from top players in the AI space, including Google, Microsoft, OpenAI and Anthropic, to discuss AI safety risks as governments step up efforts to regulate technology. We talked.
The conference released a draft policy outlining how governments will use AI, as well as spending $140 million to launch seven new AI research institutes under the National Science Foundation. It came alongside other announcements from the White House, including a solicitation for public comment.
It is not clear what exactly was discussed. The Biden administration said the meeting “underscores this responsibility and the importance of promoting responsible, credible and ethical innovation, with safeguards that reduce risks and potential harm to individuals and societies.” for emphasis,” he said.
The White House was criticized for excluding academic institutions and other nonprofits from the conference. But the administration has hinted that AI experts outside the industry will one day be given a chance to speak with officials.
According to a White House statement, “This conference is part of a broader and ongoing effort to engage with advocates, businesses, researchers, civil rights organizations, nonprofits, communities, international partners and others on critical AI issues. is.”
How much did it cost OpenAI to build ChatGPT?
OpenAI reportedly lost $540 million last year developing a viral chatbot, ChatGPT.
Training large language models is expensive. It takes thousands of GPUs and experts to build systems and datasets. OpenAI’s losses have reportedly doubled. information. It’s unclear if this figure reflects the value of the cloud computing resources donated by Microsoft.
Microsoft is the startup’s largest shareholder, investing $10 billion in 75% of OpenAI’s profits until it breaks even. ChatGPT sparked the current boom in language models and is free to use.
However, OpenAI makes money by selling API access to enterprises and offers a variety of generative AI products, including GPT-based language models, text-to-image DALL-E systems, and the Whisper model for speech generation. I’m here.
The company reportedly expects to hit $200 million in revenue this year, up from $28 million last year. Reuters previously reported that it expected OpenAI’s revenue to jump to $1 billion in 2024. OpenAI has raised $300 million worth of new funding from top VC firms including Sequoia Capital and Andreessen Horowitz, bringing its valuation to $27 billion.$29 billion according to tech crunch.
Microsoft removes waiting list for Bing AI
Netizens can now chat with Bing, Microsoft’s AI-powered web search chatbot, without being waitlisted and approved.
GPT-4 powered agents are designed to help people surf the Internet in a more conversational way. Instead of sifting through web pages to find relevant information, Bing can provide answers to queries by summarizing text. But be careful. These new Internet search chatbots do not always provide accurate or up-to-date information.
Microsoft announced the new Bing three months ago and has been improving the software to keep it on track. Bing was known to produce unstable responses, especially in its early stages. Still, people continue to flock to the site because they enjoy interacting with the new toys, according to Yusuf Mehdi, Microsoft’s corporate vice president and chief consumer marketing officer.
“Bing has grown to over 100 million daily active users and daily installs of the Bing mobile app have quadrupled since launch. It follows eight consecutive quarters of growth in Microsoft Edge browser share,” he wrote in a blog post.
Bing’s capabilities have also been enhanced to process and generate data in a variety of formats, including images and videos, in addition to text. Microsoft is working to increase usage by allowing Windows users to access the chatbot as a taskbar app. Mehdi said the move could reach more than 500 million customers each month.
US TV and film writers strike over AI
The Writers Guild of America, a trade union group representing writers working in television and film, is concerned about job losses from AI.
Large language models like ChatGPT were on the minds of their members as they announced they would go on strike this week after failing to reach agreements with top production studios over working conditions.
In addition to wages and job security, unions [PDF] Regulating the use of AI in industry. Writers want to ensure that ChatGPT cannot be used to create literary material and that their work is not used to train models. But officials from production companies like Netflix and HBO disagreed with their proposal.
“The challenge is to make sure these technologies are tools used by writers, not tools used to replace writers,” said John WGA’s negotiating committee screenwriter and author. August said. hollywood reporter. “The worry is that in the future we may see producers and executives using one of these tools to do the work that writers really have to do.”
