Two years ago, the company Anthropic was created by former OpenAI research executives. They recently launched Claude 2, an AI chatbot open to the public. Anthropic raised $750 million across the two funding rounds that began in March. As a result, their value is estimated at $4.1 billion. In May, they were among the select few companies to be invited to the White House and talk to Vice President Kamala Harris about responsible AI development.
A new entrant has emerged into the growing AI race--Anthropic. Not even a few months after it raised $750 million across two financing rounds, the startup is launching its first consumer-focused AI chatbot, Claude 2. Anthropic was founded in 2021 by former OpenAI research executives with backing from companies such as Google, Salesforce and Zoom, and it has already acquired a waitlist of over 350,000 people who are seeking access to its application programming interface and offerings. According to Anthropic co-founder Daniela Amodei, they have been focusing on making Claude as safe as possible: "We really feel that this is the safest version of Claude that we've developed so far, and so we've been very excited to get it into the hands of a wider range of both businesses and individual consumers." Initially, Claude 2 is only available to users in the U.S. and U.K., but Anthropic has plans to expand its availability in the near future.
Generative AI chatbots have been a hot topic of discussion since OpenAI rolled-out ChatGPT last year. Despite warnings from academics and ethicists about potential bias propagation, the technology quickly found its way into various sectors, including schooling, travel, medicine, advertising and more. Earlier this year, OpenAI upgraded its underlying tech with GPT-4, and allowed ChatGPT to access the internet for responses by superseding its initial training data cutoff date before disabling the chatbot's internet browsing capability as a preventative measure against misuse. Meanwhile, Google debuted a competitive Bard AI service in February and has been moving forward with enhancing its math and coding skills as well as integrating the AI into products. Anthropic is not shying away either: with Claude 2, it claims the chatbot can summarise up to 75,000 words compared to ChatGPT's 3,000 limit, and can provide summaries in the form of a memo, letter or story.-powered customer service
Daniela Amodei noted that Anthropic had invested at least two months in creating the Claude 2 AI model with a group of 30-35 individuals working directly on it and 150 people attending to it in other capacities. The marketing conditions have become so encouraging that there is considerable room for multiple players to be victorious, she added. That month, Anthropic was invited to the White House with Google, Microsoft, and OpenAI for a meeting to talk about mindful AI production. $450 million was also raised by the company in May, on the back of a $300 million funding round in March that valued the firm at a $4.1 billion price.Amodei then provided some demonstrable illustrations of the enhancements of the new chatbot compared to its predecessor. It scored 76.5% on the Bar exam's multiple choice segment, in comparison to 73%. It also obtained a 71% result in a Python coding assessment, a significant upgrade from the 56% in its former iteration.On the subject of safety, the upgraded chatbot was twice as efficient in providing “safe answers”, as reported by a blog post. Nevertheless, it is worth noting that their evaluation of safety was conducted in-house.Despite the enhancements of Claude 2 in terms of performance, Amodei acknowledged there are numerous obstacles ahead for the entire sector, including Anthropic. This comprises the issue of chatbots replacing genuine answers, known as “hallucinations” by certain technology firms.Amodei stated that “there are a lot of unknown unknowns but also known challenges for all language models today. No language model is entirely invincible to hallucinations, including Claude 2.”
Comentários