Not daring to launch a ‘homemade’ chatbot to declare war, Google is having to ‘take a detour’ to fight Microsoft

Google just invested about $300 million in artificial intelligence startup Anthropic, according to FT. The move makes it the latest tech giant to pour capital into a generation of startups trying to establish themselves in the field of artificial intelligence AI.

Under the agreement, Google will hold about 10% of the shares and make a financial investment in Anthropic. Part of the impetus comes from the fact that big tech companies need access to cloud computing platforms to process the massive AI models developed by groups like Anthropic.

Three years before Google, rival Microsoft also poured 1 billion USD into OpenAI to build a series of breakthrough AI systems, most notably ChatGPT launched late last year. It is known that this is an artificial intelligence model that can answer user questions like an entity in the form of text, stir up the media and attract 1 million users in less than 1 week of launch. .

According to the FT, both OpenAI and Anthropic are looking to develop generalized AI; Complex computer programs can write scripts and create art in seconds. While Microsoft seeks to integrate OpenAI’s technology into many services, Anthropic’s involvement with Google is limited to providing technology in the AI ​​race.

Anthropic was founded in 2021 by a group of former OpenAI researchers. Before that, they left the company out of concern that Microsoft’s first investment in OpenAI would lead to commercialization and deviate from its original focus on the safety of advanced AI.

It is known that Anthropic has developed an intelligent chatbot called Claude and competes directly with OpenAI’s ChatGPT. The startup successfully raised more than $700 million from several investors by the end of 2022, of which, Alameda Research, the crypto hedge fund of FTX founder Sam Bankman-Fried, invested up to $500 million. before filing for bankruptcy.

According to experts, Google’s move to bring Anthropic’s data-intensive computing to data centers is aimed at catching up with Microsoft, which has so far made too many strides in the booming AI market thanks to investments in technology. OpenAI. Google is also working with other startups like Cohere and C3 to try to improve their foothold in the AI ​​scene.

Previously, Microsoft CEO Nadella was said to intend to integrate OpenAI technology into the Bing search engine as well as many other Microsoft services. ChatGPT will also appear on Word, Powerpoint and all Microsoft products as a way to usurp Google’s search king.

“Microsoft is losing its appeal to young users compared to Google, especially as more and more teenagers take Google Docs for granted. So Microsoft will have to change this status quo. ChatGPT will be the game-changing ‘silver bullet’ between Google Docs and Microsoft Word,” said Dan Ives, technology analyst at Wedbush Securities.

However, until now, Google is still the most used search engine website in the world and it can take years for an app to change this.

Previously, facing the explosion of Chat GPT, Google was forced to issue ‘Red Alert’ or ‘Code Red’, with the meaning: ‘Two hands on the keyboard! Even if employees have to work overtime, skip meals and sleep, they have to solve the problem.”

“Maybe in a year or two, Google will be completely affected. AI will eliminate the search engine results page – where Google makes the most money,” shared Paul Buchheit, one of the creators of Gmail.

However, according to insiders, this red alert is just a metaphor for the threat of the ChatGPT tool. The purpose of Sundar Pichai – CEO of Google when sounding this alarm is: ‘You must treat the ChatGPT threat as seriously as Google Search is down’. All to ‘wake up’ internal Google not to forget to win.

In theory, ChatGPT is nothing new. It is developed based on the long-standing GPT3/GPT3.5 model. In reality, however, ChatGPT achieves a feat that other GPT-3-based products fail to do: user reach.

It is known that Chat GPT has become the first product in the AI ​​academic world that is accessible to a large number of ordinary users. The fact that netizens share interesting conversations with ChatGPT across network platforms makes this a very ‘hot’ topic.

Last year, Google released LaMDA (Language Model for Dialog Applications) large language model and MUM (Multitask Unified Model) multi-tasking model. Both have the same basic capabilities as ChatGPT, but the ability to reach users is not equal.

According to Forbes, Google has not yet dared to take risks with AI products due to concerns about the ethics of artificial intelligence. In addition, this corporation is also afraid of ‘shooting itself in the foot’ because in chatbots in particular or the concept of ‘conversational services’ in general, the most taboo thing is advertising – the main source of revenue for Google.

By: FT, Forbes