Ajit Karn
Google is not leaving any stone unturned as it gets ready to launch its own AI services. On Friday, news came out that Alphabet Inc. has put more than $400 million into the artificial intelligence startup Anthropic. This put the company in the spotlight. Anthropic has been testing a chat bot that could compete with OpenAI’s groundbreaking ChatGPT. And this move comes at a time when Google, which has been feeling the heat, has been making sneaky moves to get better at AI.
Dario Amodei, who was VP of research at OpenAI, and his sister Daniela started Anthropic in San Francisco in 2021. Their goal was to make large-scale AI systems that are easy to understand and reliable. TechCrunch said that Dario left OpenAI because he thought the AI industry needed to move beyond a few innovations like GPT-3. The Amodei brothers and sisters have been working on the idea that most AI models are not fully understood.
ICYMI: Google vs. Open AI: How the search giant will go up against ChatGPT
In January, the AI community found out that Claude, an AI system made by a company started by former OpenAI employees, was similar to ChatGPT. It was only available in a closed beta, so not much is known about it. In any case, several beta testers have talked about their experiences with Claude on social media.
Claude said that Anthropic made him with a method called “constitutional AI.” In a Twitter thread posted on December 16 of last year, the company said that Claude’s goal is to offer a “principle-based” approach that aligns AI systems with human goals. This means that it uses artificial intelligence (AI) like ChatGPT to answer questions based on a simple set of rules.
Anthropic started building Claude by putting together a list of 10 rules that work together to make a “constitution.” Even though these principles haven’t been made public yet, Anthropic has said that they are based on the ideas of beneficence, which means maximising good effects, nonmaleficence, which means not giving bad advice, and autonomy, which means respecting the right to choose. Claude is a statistical tool that predicts words, just like ChatGPT and other language models.
Anthropic says that Claude read a lot of text from the web and learned how words appear based on the meaning of the text around them. This has made it possible for Claude to have open-ended conversations, tell jokes, and talk about a wide range of philosophical ideas.
How does Google feel about all of this?
The investment has not been talked about by either Google or Anthropic. But they have also announced a separate partnership in which the AI startup will use Google’s cloud computing services. The investment comes at a time when ChatGPT is very popular and Google’s parent company missed both its top and bottom lines in the fourth quarter. Recent developments like ChatGPT and other AI-powered technologies seem to have pushed the search engine giant to come up with new ways to stay on top.
The search giant is worried about the growth of OpenAI’s chatbot. A normal search would give you links and images related to your questions, but ChatGPT gives you short sentences that sound like they were written by a person. The app goes a step further by letting you make content from scratch, which is a very disruptive feature. All of these things show that big changes are coming for Google and its products and services.
Also read:- Report: UAE will host the Asia Cup, which was originally assigned to Pakistan.
Using AI to make changes
Maybe this is why Google seems to be rushing to come up with its own AI breakthrough. The New York Times recently said that the tech giant has asked Larry Page and Sergey Brin, the founders of its parent company Alphabet, for help. In the same report, it was also said that Google’s founders had already agreed to plans that included adding chatbots. The company wants to release more than 20 AI products and will show off a version of Google with chatbot features.
All of this happens at the same time that Microsoft is putting a lot of money into AI and ChatGPT. Google has been working with the language model LaMDA (Language Model For Dialogue Applications). With Anthropic in the mix, it remains to be seen how powerful Google’s AI vision will be and if it will take over the AI industry.
