Claude AI Passed Law And Economics Exam, Surpassed Many Humans

A professor at Virginia’s George Mason University recently tested Claude AI on a university-level law and economics exam.

Believe it or not, this artificial intelligence earned a “marginal pass.” The professor praised the program because it produced answers “better than many human responses.”

As a result, the George Mason educator considered Claude “a competitor” and “improvement over OpenAI’s ChatGPT.

What are the advantages of Claude AI vs. ChatGPT?

Photo Credit: blog.thomasnet.com

The world has become focused on AI tools with the launch of ChatGPT last year. It has impressed people with its uncanny ability to produce convincingly manmade text.

Eventually, the phenomenon piqued the interest of other tech companies, so they began to create rivals to OpenAI’s program.

Anthropic was one of these companies. Notably, it received funding from the disgraced FTX head Sam Bankman-Fried and his romantic partner, Alameda Research CEO Caroline Ellison.

At the time of writing, it limited public access to its Claude AI and only opened closed beta testing. 

Yet, it has impressed the academic community with its ability to provide detailed answers to complex prompts. 

George Mason University Alex Tabarrok ran a law and economics exam through this program. Consequently, it earned a “marginal pass” mark and surpassed several human test takers.

Tabarrok noted one law exam question asking how to change intellectual property laws. He said it produced believable recommendations. 

Comparing Claude AI to ChatGPT revealed many of the latter’s limitations. AI company Anthropic shared a paper explaining how Claude uses “constitutional AI.”

The researchers wrote: “We experiment with methods for training a harmless AI assistant through self-improvement, without human labels identifying harmful outputs.”

Also, Anthropic tweeted, “Language models trained to be ‘harmless’ have a tendency to become useless in the face of adversarial questions.”

“Constitutional AI lets them respond to questions using a simple set of principles as a guide.” Moreover, Enterprise AI app developer Scale praised Claude AI for its “ability to write coherently about itself.”

Conclusion

Anthropic’s Claude AI recently passed university-level law and economics exams. As a result, academics showered it with praise.

As a result, it proves that it is a serious competitor for ChatGPT. It must improve quicker as more companies like Google are developing new AI programs.

More importantly, we must adapt to artificial intelligence as it transforms daily life faster. Start by following the latest digital trends at Inquirer Tech.