Anthropic’s Claude 3 AI chatbot has “near human” abilities

  • Anthropic has announced three new models within its Claude 3 lineup of generative AI offerings.
  • It now gives Claude 3 the ability to automate tasks, analyse data, power live chats, and more.
  • Anthropic also states that it outperforms OpenAI’s GPT-4 and Google’s Gemini in several key areas.

Right now the leading name in the generative AI (genAI) conversation is OpenAI. Google has entered the fray with Gemini, but OpenAI’s biggest competition may come from Anthropic, which was coincidentally spun up by a handful of former employees. To that end there startup has revealed three new models as part of its Claude 3 AI chatbot.

“Today, we’re announcing the Claude 3 model family, which sets new industry benchmarks across a wide range of cognitive tasks. The family includes three state-of-the-art models in ascending order of capability: Claude 3 Haiku, Claude 3 Sonnet, and Claude 3 Opus. Each successive model offers increasingly powerful performance, allowing users to select the optimal balance of intelligence, speed, and cost for their specific application,” Anthropic explained in a blog post.

It added that Sonnet and Opus are available now to purchase access to in 159 countries, while Haiku (featured in the video below) is expected in the coming weeks and months.

In terms of access to the models, Sonnet is available for free via an email sign-up, while Opus is available for $20 per month via a subscription to Claude Pro. It is worth pointing out too that each model will have different costs per each million tokens required for inputs and outputs, with pricing ascending or depending on the power and capability of the model itself.

Anthropic also notes that its new models outperform the likes of GPT-4 and Gemini in a number of key areas, with the Opus model in particular exhibiting, “near-human levels of comprehension and fluency on complex tasks.”

“All Claude 3 models show increased capabilities in analysis and forecasting, nuanced content creation, code generation, and conversing in non-English languages like Spanish, Japanese, and French,” Anthropic continued.

While Haiku’s availability is still in the offing, Anthropic has noted that it is, “the fastest and most cost-effective model on the market for its intelligence category. It can read an information and data dense research paper on arXiv (~10k tokens) with charts and graphs in less than three seconds.”

For those worrying about inaccuracies, or hallucinations, of generated content, the startup highlighted that, “compared to Claude 2.1, Opus demonstrates a twofold improvement in accuracy (or correct answers) on these challenging open-ended questions while also exhibiting reduced levels of incorrect answers.”

It remains to be seen whether Claude 3 generates the fervour that ChatGPT has, but it’s clear that Anthropic’s immediate concerns for its genAI models is beating what OpenAI makes available.

As for who is regulating the pace at which these solutions are being rolled responsibly, is still unclear.


About Author


Related News