Anthropic launched a significant upgrade to its conversational AI assistant, Claude 2, with 100k context windows, and the ability to upload five files in a chat.
The new Claude-2-100k model provides improved performance, more extended responses, and expanded access through a public beta website and API integration.
Passing AI Detection: Claude 2 Vs ChatGPT
The content generated by Claude 2 is considered more human-like by AI detection tools than content from ChatGPT.
I tested several pieces of content generated by both AI chatbots.
The following experiment was with the prompt: Can you write a 500-word article about SEO?
Originality.ai’s detection tool was 98% confident that a human created Claude 2’s content.
However, the plagiarism detector did note that the first sentence from Claude 2’s response was pulled directly from an article.
ZeroGPT deemed Claude 2’s content as mostly human written.
ChatGPT’s response had zero plagiarism but was identified as 100% likely to be AI-generated.
ZeroGPT also deemed ChatGPT’s content as AI-generated.
It’s important to note that while Claude 2 may produce more human-like content, you will still want to give your prompts unique elements to ensure that you don’t publish a piece of content that matches what someone else generated with the same tool and the same basic prompt.
Originality and other AI detection tools have since been updated and now can detect Claude 2-generated content as AI.
Claude 2 Analyzes Up To Five Files At A Time
Another useful feature: uploading up to five documents for analysis in a chat.
The new model of Claude can handle longer input and output, analyzing documents of up to 100,000 tokens and writing content with up to a few thousand tokens.
I tested the upload feature in Claude 2 with a 31-page PDF file of a study, and it quickly responded with the following.
I also asked it some basic questions about an exported GA4 report.
This makes it the perfect alternative to any ChatGPT plugin offering PDF analysis.
Claude 2 Has Improved Reasoning Capabilities
Anthropic’s new release demonstrates stronger language and reasoning capabilities than previous versions, scoring higher on tests like the BAR exam and GRE.
Introducing Claude 2! Our latest model has improved performance in coding, math and reasoning. It can produce longer responses, and is available in a new public-facing beta website at https://t.co/uLbS2JNczH in the US and UK. pic.twitter.com/jSkvbXnqLd
— Anthropic (@AnthropicAI) July 11, 2023
The company claims this model has much better coding abilities, with significantly higher scores on programming evaluations.
Anthropic says safety has been a focus in developing Claude, making it less likely to generate harmful content. The company uses extensive techniques and testing to reduce risks.
The public website offers users disclaimers about the AI assistant upon signup about the reliability of the information.
It also lets users review the company’s safety policies.
How To Get Claude 2
Claude-2-100k is available in a public beta for users in the US and UK through the new website and Poe.com with a paid subscription.
Today we are launching new, more powerful models with increased context windows on Poe, including the just-released Claude 2 with its 100k-token window length, along with a set of new tools to help everyone take advantage of them. (thread below) pic.twitter.com/iJAC066wRQ
— Poe (@poe_platform) July 11, 2023
Claude API Early Access For Businesses
The Claude API is also offered for businesses, with the same pricing as the previous version.
You can sign up for early access to the Claude API here: https://www.anthropic.com/earlyaccess
Early partners, including Jasper and Sourcegraph, are already integrating the new version of Claude via the API and finding value in its enhanced semantics, reasoning, and larger context window.
Antrhopic’s news coincides with Jasper’s layoff announcement to focus on AI:
And so today, with a heavy heart, I told the team at Jasper that we will be discontinuing a number of roles so that we may sharpen our focus and align resources to become the best possible AI copilot for marketing teams.
The Future Of Conversational AI Has Arrived
With the rapid advancement of conversational AI, Anthropic is becoming a reliable alternative to ChatGPT and other generative AI chatbots.
Featured image: Tada Images/Shutterstock