[ad_1]
Claude 2 from Anthropic is a powerful new AI tool that marketers should pay attention to.
Anthropic, an artificial intelligence major, has just announced the release of Claude 2, its new, more powerful AI foundation model.
And marketers should take note, as the enhanced capabilities of Claude 2 open up some seriously impressive uses for marketing and business.
“It’s pretty significant,” says Paul Roetzer, founder/CEO of the Marketing AI Institute. “I’d say it’s probably the closest to GPT-4 that I’ve personally experienced.”
Here’s what you need to know.
What is Claude 2?
Marketing AI Institute founder/CEO Paul Roetzer and I discussed Claude 2 in detail in episode 55 of The Marketing AI Show.
First, what is it?
Claude 2 is a large language model developed by Anthropic. Like ChatGPT, Claude 2 responds to your natural language requests. You can use it to generate text, answer questions, generate code, and much more.
It also has much stronger reasoning capabilities in math and code than its predecessor. (He scored above the 90th percentile on the GRE reading and writing test.)
But perhaps most importantly, Claude 2 now has a 100,000 token context window. Tokens are individual parts of words that are processed by a Claude-like model. For context, the most powerful version of GPT-4 has a 32,000-token context window. The standard version has an 8000-token window.
This means you can comfortably upload an entire novel The Great Gatsby Claude 2, which weighs around 72,000 tokens, with a spare.
According to Anthropic: “This means that Claude can run hundreds of pages of technical documentation or even a book.”
In fact, Claude 2 makes it easy to cut and paste a lot of text, upload PDFs, and upload CSVs, so you can use the model for a huge amount of material for different use cases.
How does the Claude 2 perform against other models?
To test the Claude 2, our CEO, Paul, put it through its paces against other models in a very specific and relevant use case.
As part of a chat at this year’s Marketing Artificial Intelligence Conference (MAICON), Paul interviews AI expert Ethan Molick. In preparation for this interview, Paul enlisted the help of Claude 2, Google Bard, Inflection’s Pi, ChatGPT using GPT-3.5, and ChatGPT using GPT-4 with Code Interpreter.
He gave each tool a series of articles written by Mollick. For each article, Paul asked for a summary of each article and suggestions for 5-7 potential questions to ask Mollik.
By far the best performers were Claude 2 and GPT-4.
“These summaries were incredible and the questions were profound,” says Paul. It also indicates that GPT-3.5 has also done a solid job.
Surprisingly, Bard was relatively poor at this particular task. He would summarize the articles, but even if he had access to the Internet, he would miss obvious or important details.
Inflection’s Pi also had problems. Pi should be a more talkative and human-like agent than the others. But this had its downsides. Paul said Pi kept asking him questions over and over, making comments and using emotions to humanize the conversation, but he didn’t actually complete the task.
Paul ended up taking the results from the Claude 2 and GPT-4, as well as his own questions, and combined them into a final list of 20-30 questions for the interview. About half of the intended questions came from two instruments.
In some cases, these two tools framed the questions better than Paul and were given additional context rather than Mollick’s articles to create deeper, more useful summaries.
Should Marketers Use Claude 2?
“I know there are a lot of competing products out there, but I think it’s worth paying attention to,” says Paul. If you haven’t, you should add Claude 2 to your list, especially given its large context window.
But, Paul says, you have to keep coming back all from the main tools. They change dramatically and improve in a short period of time. Just because some of them failed this challenge doesn’t mean they won’t do well in the future.
Paul recommends having 3-5 standard use cases for which you use these types of tools. Think: transcribing, summarizing, writing first drafts, etc. Every month or two, try these use cases across multiple tools and models because they’re in a constant race, with different models moving forward at different times.
Don’t be left behind…
You can stay ahead of AI-driven disruption—and fast—with us Piloting AI for Marketers course seriesA series of 17 on-demand courses designed as a step-by-step learning journey for marketers and business leaders to increase productivity and performance through AI.
The course series includes 7+ hours of learning, dozens of AI use cases and vendors, a collection of templates, course quizzes, a final exam, and a professional certificate upon completion.
After getting Piloting AI for Marketers, you’ll:
- Learn how to advance your career and transform your business with AI.
- Have 100+ use cases for AI in marketing — and learn how to identify and prioritize your own use cases.
- Discover 70+ AI vendors across various marketing categories that you can start piloting today.
[ad_2]
Source link