Anthropic, an OpenAI rival, opens Claude 2 AI chatbot to the public

There is a new entrant in the emerging AI arms race.

like MicrosoftSupports OpenAI and Google In a race to develop more advanced chatbots, powered by generative AI, Anthropic is investing heavily to keep up. Just a few months after raising $750 million in two funding rounds, the startup is launching a new AI chatbot: Cloud 2.

Founded in 2021 by former OpenAI research executives and funded by companies including Google, sales force And Zoom inAnthropic opens chatbot technology to consumers for the first time with Claude 2. Over the past two months, the company’s AI models have been tested by companies like Slack, Notion, and Quora, and Anthropic has amassed a waiting list of more than 350,000 people requesting access to its Cloud API and consumer offerings.

“We focused on the business, on making Claude as safe as possible,” said Daniela Amoudi, who co-founded Anthropic with her brother Dario. “We really feel this is the most secure version of Claude we’ve developed to date, and so we’ve been very excited to get it in the hands of a wide range of businesses and individual consumers.”

Cloud 2 will initially only be available to users in the US and UK, and Anthropic plans to expand availability in the coming months.

Since OpenAI introduced ChatGPT to the public late last year, the tech world has invested heavily in the capabilities of AI-powered chatbots, which respond to text prompts with complex dialogue responses. Academics and ethicists have expressed great concerns about technology’s tendency to spread bias, but even so, it has quickly made its way into schools, online travel, the medical industry, online advertising and more.

In March, OpenAI released GPT-4, its biggest update to the underlying technology behind ChatGPT. After 2 months ChatGPT was allowed to start surfing the internet so responses were no longer restricted to the 2021 training data cut off date, but then Disable the chatbot’s internet browsing ability Yet reports of problematic uses such as bypassing paywalls.

Google announced its rival Bard in February, and has since updated the chatbot’s math and coding skills and integrated the AI ​​service into more products.

Anthropic aspirations are no less.

The company says Claude 2 has the ability to summarize up to 75,000 words, which can be as long as a book. Users can enter large groups of data, and request summaries in the form of a note, letter or story. By contrast, ChatGPT can handle about 3,000 words.

Daniela Amoudi said Anthropic has invested at least two months in developing Cloud 2, with a team of 30-35 people working directly on the AI ​​model and a total of 150 people supporting it. She said the market is growing so fast that there is plenty of room for many players to find success.

“It’s really an unusual time from a business perspective because there’s a very high demand for large language models and more demand than the industry can currently provide,” Amoudi said. “The landscape is very broad, and there is really very much room for many different users and different types of users to take advantage of these systems.”

In May, Anthropic was one of four companies invited to a meeting at the White House to discuss responsible AI development with Vice President Kamala Harris. Google’s parent company Alphabet, Microsoft and OpenAI were the others.

In the same month, the company raised $450 million, following a $300 million funding round in March at a valuation of $4.1 billion.

Amodei provided some concrete examples of Claude 2’s improvements over the previous version of the model. The new chatbot scored 76.5% on the multiple-choice section of the bar exam, up from 73%. And in the Python coding test, Cloud 2 scored 71%, up from 56% in the previous iteration.

When it comes to safety, the upgraded chatbot was twice as good at giving “harmless responses,” according to a blog post though it’s worth noting that the Red Team’s safety assessment was done internally.

Despite Claude 2’s performance improvements, Amodei acknowledged that there are a lot of hurdles for Anthropic and the entire industry. For example, the tendency of AI chatbots to fabricate incorrect answers, called “hallucinations” by some tech companies, is an ongoing problem.

“There are a lot of unknowns left but there are also known challenges with all paradigms of language in the world today,” Amoudi said. “No language model is 100% immune to hallucinations, and Claude 2 is the same.”

He watches: ChatGPT boosted optimism about AI