For two weeks, I decided to do something I wasn’t sure I wanted to do: I cheated on ChatGPT.

Well…sort of.

I opened up Claude.

I’ve been using ChatGPT for a long time, and one of the reasons I keep coming back to it is because I’ve spent a lot of time teaching it what I need.

It knows how I like my stories written. It knows the kind of HTML I want. It knows how I structure my articles, the type of SEO information I want, and, sometimes, it even sounds like me.

So when I started hearing people rave about Claude, my first thought was:

Do I really want to start all over again with another AI?

But curiosity won.

I decided to open the free version of Claude and actually use it for two weeks instead of just reading other people’s opinions about it.

And I learned something interesting.

I don’t think there is one AI that does everything best.

Claude Was Easy to Get Used To

Aesthetically, I liked Claude immediately.

If you’re already familiar with ChatGPT, Claude doesn’t feel like you’re learning an entirely foreign program. The interface is clean, simple and easy to figure out.

That made it easy for me to jump right into testing it.

The real question was whether it could actually handle the work I do.

I Usually Use ChatGPT for Small HTML Jobs. Claude Made Me Go Bigger.

One of the things I regularly use ChatGPT for is HTML.

I’ll ask for a little box for a story, a poll, a related-story section, a button, or some other piece of code that I can drop into my website.

With Claude, I decided to push it further.

Instead of asking for one little HTML component at a time, I gave it a much larger website-design task.

Claude actually did the entire thing.

It wasn’t perfect.

There were a couple of times when it messed something up, and I had to tell it what needed to be corrected.

But honestly, I was okay with that.

It was able to take a much bigger assignment and work through it, which made me realize Claude might be particularly useful for people who want AI to handle larger chunks of a project rather than just individual tasks.

And that got me thinking about something else.

What happens when you take that ability into the workplace?

Because website design is one thing.

But lawyers, financial professionals, researchers, and other industries are dealing with documents and information that can be hundreds—or thousands—of pages long.

That’s where Claude gets particularly interesting.

AI At Work
What Can AI Actually Do for Different Industries?
From legal research to recruiting and software development, AI is moving far beyond simple chat.
Industry What AI Can Help With Why It Matters
⚖️ Legal Contract review, legal research, document comparison, due diligence, regulatory research, deposition summaries and litigation preparation. Lawyers can spend less time sorting through massive amounts of paperwork and more time applying professional judgment.
💰 Finance Financial documents, research, reports, data analysis, summaries, presentations and repetitive administrative tasks. Professionals can quickly organize and analyze large amounts of information.
🏥 Healthcare Administrative documentation, information summaries, research support, drafting and organizing complex information. AI can help reduce time spent on information-heavy administrative work, while qualified professionals remain responsible for decisions.
💻 Software Writing and debugging code, understanding existing code, documentation, testing and building applications. Developers can use AI as a coding partner instead of starting every task from scratch.
👩‍💼 Recruiting & HR Job descriptions, resume summaries, interview questions, candidate organization and employee communications. AI can take some of the repetitive information-processing work off recruiters’ desks.
🎓 Education Lesson planning, educational materials, summaries, brainstorming, tutoring support and adapting content. Teachers and students can use AI to quickly create and reorganize learning materials.
🔬 Research Summarizing papers, organizing research, comparing information, brainstorming and working through large collections of documents. Researchers can spend less time manually sorting information and more time interpreting it.
📰 Media & Content Articles, editing, SEO ideas, HTML, research, social content, brainstorming and content planning. Writers and publishers can use AI for everything from brainstorming to technical website tasks.
🏢 Business Reports, presentations, research, meeting summaries, documents, planning and repetitive workflows. AI can become a general-purpose assistant across multiple departments.
Important: AI can assist with these tasks, but professional review and human judgment remain important, particularly in areas such as law, healthcare and finance.

Claude and the 100-Page PDF Test

One of my favorite Claude experiments involved a PDF that was more than 100 pages long.

I needed to change dates throughout the document.

Claude handled it surprisingly well.

That was probably one of the moments when I understood why people are excited about AI for professional work.

I wasn’t asking it to write a cute social media caption.

I was giving it a huge document and a tedious job.

And it could work through it.

That experience also made me realize that the AI competition isn’t really just about who writes the better paragraph anymore.

It’s increasingly about:

Who can handle the bigger, more complicated job?

And Lawyers Are Already Using AI This Way

This is where my little experiment starts connecting to something much bigger.

Anthropic says legal teams are using Claude for things like contract review, redlining, legal research, M&A due diligence, regulatory and compliance work and litigation preparation. Claude can also connect with legal software and document-management systems used by law firms.

That is a very different use case from me asking an AI to make a cute HTML box.

For example, Claude’s legal tools can compare contracts, identify changes, flag potential risks and help organize large collections of documents for review. Anthropic says its legal workflows can also help with deposition preparation and drafting memos based on multiple documents.

And that’s where my 100-page PDF experience suddenly doesn’t seem so unusual.

If AI can help one person work through a giant document, imagine what happens when a law firm connects it to an entire collection of contracts, pleadings, prior versions and case materials.

Of course, that doesn’t mean lawyers should let AI make the final legal decisions.

Claude’s own legal tools emphasize that AI-generated work needs attorney review.

That’s an important distinction.

AI can help do the heavy lifting. The lawyer still has to use judgment.

That’s Actually the Part I Find Most Interesting

The more I experimented with Claude, the more I started thinking about AI differently.

At first, I was comparing:

Claude vs. ChatGPT.

But maybe the better question is:

What can AI actually do for different professions?

Anthropic’s customer directory now lists Claude use across industries including financial services, healthcare, legal, insurance, recruiting, education, scientific research, software, transportation and professional services.

And the potential applications are pretty broad.

For lawyers

AI can help with:

  • Contract review
  • Comparing documents
  • Legal research (always check for accuracy)
  • M&A due diligence
  • Regulatory research
  • Litigation preparation
  • Summarizing large records
  • Drafting memos
  • Reviewing agreements

For recruiters and HR professionals

AI can help organize candidate information, draft job descriptions, summarize resumes, and create interview materials.

For researchers

AI can help digest lengthy papers, organize research, identify themes, and work through enormous amounts of information, but make sure you still read everything that AI provides for your research.

For software developers

AI can help write and debug code, understand existing codebases and create documentation.

For businesses

It can help with reports, research, documents, presentations, analysis and repetitive administrative work.

And that’s just scratching the surface.

My Media Kit Was a Different Story

Of course, Claude didn’t win every round.

I tried using it for a media kit, and that experience wasn’t nearly as impressive.

I also tried Meta AI for some of this type of work, and neither one really became my favorite for that particular task.

That’s when I started realizing something important:

Just because an AI is great at one thing doesn’t mean it’s great at everything.

And that’s probably going to be one of the biggest lessons for people using AI over the next few years.

We may stop asking, “What’s the best AI?”

And start asking:

“What’s the best AI for this job?”

ChatGPT Still Wins for Images—For Me

Images were another area where I preferred ChatGPT.

I didn’t really like the way Claude handled pictures, so I didn’t end up using it much for image creation.

I already use Meta AI for a lot of my photos, but when I’m comparing it with ChatGPT, I prefer ChatGPT’s image results.

For me, there’s more variety, and I generally prefer the realism.

That matters because images are a huge part of what I do and need as an independent Journalist that can’t always afford Getty Images, etc.

So for image generation, ChatGPT remains my pick.

But There Is One Big Catch: I’m Using the Free Versions

This entire experiment comes with an important disclaimer.

I’m comparing the free versions.

That means I’m not necessarily experiencing everything these companies offer to businesses paying for enterprise-level products.

And there are limits.

With both ChatGPT and Claude, you can only do so much with the free versions before you start running into usage restrictions.

That can become especially noticeable when you’re using AI heavily for writing, coding, documents, images, and research.

So I’m not going to pretend my two-week experiment proves that one company is objectively better than the other.

It doesn’t.

It tells me which tool works better for me.

So, Am I Leaving ChatGPT?

No.

But I’m also not giving Claude back.

That’s probably the biggest surprise.

ChatGPT is still my primary AI.

I’ve spent too much time building a workflow around it to simply walk away.

And there’s something else that matters: familiarity.

When you’ve used a tool for a long time, you know how to talk to it.

You know what prompts work.

You know what it can handle.

You know how to fix it when it gets something wrong.

That has real value.

But Claude earned a place in my toolbox.

If I have a massive document to work through, I’ll think about Claude.

If I’m doing a larger website-design project, I’ll consider Claude.

If I’m working with certain types of long-form documents, I now know it can be surprisingly useful.

For images?

I’m sticking with ChatGPT and Meta AI.

Maybe AI Isn’t About Picking a Winner

After two weeks, I’ve come to a conclusion I didn’t expect.

I don’t think I need to choose one.

Why would I?

If Claude is better for one job, I’ll use Claude.

If ChatGPT is better for another, I’ll use ChatGPT.

If Meta AI gives me the image I need, I’ll use Meta AI.

It’s like having different tools in a toolbox.

You wouldn’t use a hammer to tighten a screw just because the hammer happens to be your favorite tool.

The same thing is happening with AI.

And the professional world appears to be moving in the same direction.

Anthropic says legal teams are already using Claude for document-heavy workflows such as contract review, due diligence, regulatory work and litigation preparation, while its customer examples span numerous other industries.

That makes my little experiment with a 100-page PDF seem pretty small.

But maybe that’s exactly the point.

The AI revolution isn’t just about writing blog posts or making funny pictures anymore.

It’s about taking tasks that used to consume hours of someone’s day—and asking whether a machine can help get them done faster.

My Two-Week Verdict

ChatGPT: Still my overall favorite. It fits my existing workflow, understands what I need, and remains my favorite for images.

Claude: A surprisingly strong competitor, especially for large documents and bigger website-design projects.

Meta AI: Still useful for images and photos and has a place in my workflow.

The winner: Honestly, there isn’t one.

The best AI might simply be the one that does the job you need done best.

And since I’m using the free versions?

I’m going to keep doing what works.

I’ll just get in where I fit in.


Apple MacBook Neo Laptop

Apple MacBook Neo

Price: $589.99

The MacBook Neo offers an affordable entry into Apple’s ecosystem with a lightweight design, long battery life, and enough performance for note-taking, web browsing, streaming classes, writing papers, and everyday productivity. It’s an excellent choice for students looking for a budget-friendly Mac laptop.

Check Price on Amazon

As an Amazon Associate, My Unbounded Life may earn from qualifying purchases.