ChatGPT has become a major disruptor in the technology world, with the artificial intelligence tool being viewed as a potential replacement for Google search. Based on its capabilities, Microsoft has already announced the inclusion of the ChatGPT AI model in the Bing search engine. Google appears to be looking for a counter, as it has been reported that its parent company Alphabet inc. has invested $300 million in a AI systems company Anthropic, which was founded by ex-OpenAI employees.
Anthropic is working on Claude, an AI assistant tool that could compete with OpenAI's ChatGPT. According to the Financial Times, Google has invested around $300 million for a 10% stake in the startup.
All you need to know about ChatGPT's potential rival Anthropic
1. Anthropic is an AI safety and research firm that claims to be working on building reliable, interpretable, and steerable AI systems. It aims to build a model that is not 'unpredictable, unreliable, and opaque'.
2. In 2021, Dario Amodei, former vice president of research at OpenAI, and his sister Daniela founded Anthropic in San Francisco. ALSO READ: Loving ChatGPT? Check these other 5 AI tools
3. Anthropic created the AI assistant Claude using safety techniques such as Constitutional AI. According to the company, the AI model has been trained to be a harmless AI assistant through self-improvement, with no human labels identifying harmful outputs. The only human oversight is provided through a list of rules or principles, and it is thus referd as 'Constitutional AI'. This enables the development of a non-evasive but harmless AI assistant that engages with harmful queries by explaining its objections to them. ALSO READ: This AI tool from Google can generate music from text
{{/usCountry}}3. Anthropic created the AI assistant Claude using safety techniques such as Constitutional AI. According to the company, the AI model has been trained to be a harmless AI assistant through self-improvement, with no human labels identifying harmful outputs. The only human oversight is provided through a list of rules or principles, and it is thus referd as 'Constitutional AI'. This enables the development of a non-evasive but harmless AI assistant that engages with harmful queries by explaining its objections to them. ALSO READ: This AI tool from Google can generate music from text
{{/usCountry}}4. Anthropic trains, scales, and deploys its AI systems on Google Cloud's GPU and TPU clusters.
5. It is unclear whether Google will incorporate Claude into its services, as Microsoft has done with ChatGPT, or if the deal is merely a means of expanding Google's cloud computing business.