ChatGPT 3.5 vs. Claude - An Overview of A GPT Alternative While You’re On the Waitlist

Many are eager to try out ChatGPT 4, the next version of OpenAI's popular conversational AI system. However, accessibility is currently limited and the waitlist for this new version is over 3 months long. For those of us who have been exploring the integration of AI in our work flows, this proves challenging when the 3.5 version just isn’t cutting it. Fortunately, I was able to stumble across a promising alternative, Anthropic’s, after reading about it in our company-wide AI Slack Channel. 

In this post, we'll explore Claude and my experience using it on a task that ChatGPT 3.5 wasn’t able to process.

What Is Anthropic? 

Anthropic came into existence in 2021, thanks to a group of former high-ranking members from OpenAI. The Amodei siblings, who founded Anthropic, were actually part of the crew that left OpenAI back in 2019 due to differing POVs on partnering with Microsoft.  Anthropic has also recently partnered with Amazon as its primary cloud provider.

What Is Claude AI?

Claude is an artificial intelligence assistant created by Anthropic that specializes in augmenting human capabilities through reliable functionality and transparent constraints. Its Constitutional AI training methodology works to align Claude’s behaviors with human values. Rather than speculating, Claude deliberately avoids potential harms by admitting knowledge gaps, refusing unethical directives, explaining its reasoning, and allowing corrections when mistakes happen (Drapkin).

What Can Claude Do?

Read, analyze, and summarize uploaded files

When searching for a way to quickly map keywords out at the ad group level for a client, I tested out using Claude. One major pro for me was being able to upload files to Claude for easy access. Files that are uploaded to Claude can be Word documents, PDFs, .txt, .csv, and others. It doesn't handle Excel spreadsheets but changing your file type should solve that pretty quickly.  

Process more data than ChatGPT.

 When uploading a list of keywords in a campaign, with performance data. ChatGPT 3.5 was only able to handle 100 of my 482 active keywords while Claude was able to accept a file of all my keywords with their performance over time. A major pro if you’re a habitual spreadsheet-opener like me.

Access Links (Kinda)

When utilizing Claude while mapping out negative keyword recommendations, one common theme I got across both ChatGPT 3.5 and Claude was that AI kept recommending negative keywords that were going to conflict with the keywords I’m actively bidding on. When I attempted to link a Google Support article to Claude so it could review its negative keyword recommendations, I immediately received an AI wheel of despair. However, I was pleasantly surprised when it provided a semi accurate summary of the article immediately despite the warning.


Consider Claude if You’re In Need of Something More out of Your AI Adventures. 

I’m always on the lookout for ways to incorporate AI to make my life easier at work. I’m confident if you’ve made it to the end of this blog, you are too. When using ChatGPT 3.5 and Claude for the same task, I found the experience easier and more accurate utilizing Claude, will continue to work it into my day-to-day while I’m on the waitlist. 

We love helping marketers like you.

Sign up for our newsletter to receive updates and more:

Jess Henderson
Jess Henderson
Sr. Associate, Paid Media