What Is Claude AI? A Close Rival to OpenAI’s ChatGPT
Claude AI is one of the closest competitors to ChatGPT. It was developed by Anthropic, a startup founded by former OpenAI employees Dario Amodei and his sister Daniela Amodei. Claude was trained using a set of principles referred to as Constitutional AI, to be safe, accurate, and secure.
Anthropic’s bot stirred the world of generative AI, receiving billions of investments from Amazon, Google, and other companies. The startup was founded in 2021 and released Claude’s first version in March 2023. Several months ago, it hit a valuation of $15 billion, according to the New York Times.
Anthropic has introduced several Claude models since its launch. The latest iteration was the Claude 3 family, released on March 20, this year. It consists of the Claude 3 Opus, the Claude 3 Sonnet, and the Claude 3 Haiku. Among them, the Opus is the most intelligent and powerful version. It is available only for Claude Pro subscribers and costs $20 per month. Opus is suitable for complicated tasks, including advanced research and automation.
Sonnet is Claude’s mid-range version that powers its free version. To use it, you just need to open an account through your email. Then you start a chat and tell the bot how you want it to assist you. For example, you can attach a document and instruct the bot to summarise it, ask questions that interest you, or generate code.
Finally, Haiku, Claude’s third version is the most compact and fastest one. Due to its responsiveness, Haiku can be used in customer support, data processing, and other applications.
Claude's Strengths and Capabilities
The first thing that you’ll probably notice about Claude is its human-like language. The answers of the bot are friendly and natural. It can provide information up to August 2023 on a broad range of topics.The bot's knowledge, however, is cut off, as it can't browse the web for new information, which puts it at a disadvantage compared to competitors.
Using Claude comes with a daily message limit, which depends on the bot’s demand. The system notifies you when you’re close to hitting the limit.
Claude excels at reasoning, mathematical problem-solving, and programming, among other tasks. Because the system focuses on safety and ethics, it is highly unlikely to provide harmful responses. Anthropic has been using the methodology of filtering out bias and discrimination from the model. This makes Claude a good option for enterprise use. Now, the AI assistant is available in the paid version of Slack. It can remember threads, analyze content from a shared website, summarize conversations, and more.
Users highly rate Claude's capabilities. On the LLM evaluation platform, LMSYS Chatbot Arena, the Claude 3 Opus is among the top three AI models, coming after ChatGPT 4-o, and Gemini 1.5 Pro by Google. Not long ago, in March, the Claude 3 Opus was the leader of the board. The rankings changed after Google and OpenAI launched their new models.
Anthropic is expected to launch a new model, Claude 4, later this year. Meanwhile, OpenAI's GPT-5 is on its way, promising increased competition between generative AI chatbots.
The Strange Consequences of Being Overly Ethical
Ethical limitations have their toll on Claude. Sometimes the bot refuses to answer questions, citing being uncomfortable. The high moral principles may require a special approach to the program when you need to start a new chat and try a renewed, more specific way to put your question. This happens in different situations, especially when requests are unusual, or relate to sensitive topics such as politics and religion. Interestingly, over time, Claude may refuse to answer the types of questions it did before.
A Reddit user, for example, shared a case when he asked Claude to write a letter in Taiwanese for his girlfriend. Despite providing the text before, Claude’s answer was that it doesn’t speak the language.
Source: www.reddit.com
In such cases, workarounds of original prompts, more context, and rephrasing may be helpful. To write a letter in a foreign language, you need to ask Claude to write it first in English and then translate it, or, explain more precisely what you need, using clear instructions.
Growing Diversity in the Generative AI Bots Space
With the booming sector of AI bots, users now have a broad range of options to choose from. They can experiment with tools, find which ones work best for them, or use multiple tools in a complementary way. In fact, it has become hard to catch up with recent developments: frequent model updates, improved capabilities, and new apps make headlines day by day. In the space, Claude, with its reasoning capabilities and high ethics, is one of the most promising AI assistants to look out for.
Anahit Avetisyan
Author
Web3 writer and crypto HODLer with a keen interest in market trends and recent technologies.
