Claude Pro: First impressions of Anthropic’s paid AI chatbot

Artificial intelligence is all over the news right now with information, narratives and opinions plentiful if not bountiful on the pros and cons of AI.

The hottest topic is generative AI, the subset of AI that, broadly speaking, is a form of machine learning that’s able to produce text, video, images, and other types of content based on user-given text prompts or dialogue. Essentially, you chat with a bot. ChatGPT and DALL-E are examples of generative AI applications that produce text or images respectively.

Last month, I published an article that compared four of the primary gen-AI applications that can create text content in response to prompts: ChatGPT, Claude, Bing, and Bard.

I mentioned that ChatGPT Plus, the paid version of ChatGPT ($24 a month including 20% UK VAT), is the app I use most with Claude AI a close second. Last week, Anthropic – Claude’s creator – introduced Claude Pro, its new paid version ($20/£18 a month subscription including 20% UK VAT) in the UK and US that changes the game somewhat in terms of capabilities compared to the free version, and in direct competition with ChatGPT Plus.

While making a comparison between the paid versions of ChatGPT and Claude would be a logical step, I’m more interested for now in a comparison between the free and paid versions of Claude to help me answer the question – is it worth the upgrade?

Well, I’ve signed up for a paid subscription to see. There’s no long-term commitment; I can cancel after a month if I wish to.

Let’s take a look at the two versions with the help of – you guessed it – Claude Pro itself.

Claudes compared

To get straight to the point, I asked Claude Pro to tell me the primary advantage and benefit of paid Claude Pro over free Claude.

If I had to highlight the primary advantage and benefit of Claude Pro compared to the free Claude (it replied), I would say it is the ability to have natural, unlimited conversations with long-term memory and personalization.

The key differences that enable this are:

  • No limits on the number of chat sessions or messages with Claude Pro, allowing for ongoing conversations.
  • Claude Pro remembers context from previous chats and your interests/preferences, allowing for more personalized, coherent dialogue.
  • More advanced AI capabilities in Claude Pro lead to deeper conversations that can understand more complexity.

Essentially, Claude Pro removes the barriers around length, memory, and complexity that exist with the free version. This allows you to have free-flowing, human-like chats without feeling like they are constantly having to restart and you’d have to re-explain topics.

The unlimited, personalized conversations enabled by Claude Pro facilitate the development of a relationship with the assistant over time. It’s able to learn about you and have discussions that feel more natural as a result. That persistent, individualized experience is the primary differentiator from the limited small talk offered by the free Claude.

(I’d add an additional comment to that point – ‘limited small talk’ is essentially what you get with the other chatbots mentioned earlier.)

So in summary, Claude Pro’s primary advantage is enabling unlimited, natural conversations with an AI that gets to know you. That personalized approach is designed to unlock the full potential for chatbot assistance.

Let’s see the key differences between paid and free Claude in a side-by-side comparison:

FeatureClaude ProFree Claude
Chat LimitsUnlimited sessions30 messages per day
Conversation ContextRemembers context and conversationsNo memory – starts fresh
AI CapabilitiesAdvanced AI and knowledge baseBasic AI only
AdsNo adsOccasional ads
New FeaturesPriority early accessDelayed access
PersonalizationCustom recommendations based on youGeneric responses
Availability24/7 with fast responseSlower responses, occasional downtime
Cross-deviceSyncs conversations across devicesIsolated conversations
SupportDirect customer supportNo support

While all this is great, there are two standouts for me:

  1. The free Claude starts each chat session fresh with no memory. But Claude Pro can remember the context and previous conversations, allowing it to have more natural, back-and-forth dialogues with you. This was immediately apparent to me in the back-and-forth with Claude Pro in researching and preparing the content for this article.
  2. Sharing information across different devices is very useful indeed as, for example, it will be immensely convenient to start a chat with Claude Pro on one device and continue it in Claude Pro on another.

I especially like the option to install a Claude Pro app on my Samsung smartphone (and tablet) as Claude told me there is one for Android and listed some compelling advantages of the app and its functionality.

And this is where we have a fork in the road that separates credibility from trust.

What Claude told me wasn’t actually true. There is no app yet, it admitted, when I told it I had searched high and low on Google Play with no such app to be found.

Claude Pro admist it made "an incorrect assumptionm".

Is this a “hallucination” of the type you hear about where chatbots make stuff up, often including completely irrelevant and even wholly untrue information, that they can’t answer?

While I don’t think it is in this case, I’m in two minds about this. Claude Pro didn’t have information that would enable it to answer my question so it “made an incorrect assumption” but based on some actual information. Is that the same as “making stuff up”? Is it egregious enough to distrust anything it says? As it noted:

Thank you for catching my mistake! I should not have provided a link or details about an Android app that does not exist yet. As an AI assistant, I do not have direct access to Anthropic’s release plans and should not speculate on unreleased products.

Claude Pro response

Here’s how I see it.

While I do believe that Claude Pro is a highly useful tool and one that I will likely use frequently (and most certainly will do during the first month of my paid sub), this illustrates two fundamental aspects:

  • Never accept at face value anything a chatbot tells you in response to your prompting.
  • Always verify everything a chatbot tells you.

That means you will need to spend time fact-checking everything, even if some of the things you’ll check are what you believe you already know. Check it anyway. And check the sources you find. (If you’re a volunteer Wikipedia editor, you’ll know all about verification, neutral point of view, and credible sources and what those terms mean.)

This is something I have done since my first experience with a generative AI chatbot (ChatGPT) earlier this year, just eight short months ago. In today’s sorry climate of such mistrust, it’s an essential part of your diligence with everything you create that you use in your work or tells a story of some kind.

Finally, there is one other consideration you must take into account.

Unlike ChatGPT Pro and its plugins, both Claude Pro and free Claude cannot research anything online. So if you prompt either version to, say, summarise an article in a publicly-accessible magazine at such-and-such a URL, you’ll get the response:

Claude can’t visit links. Responses may contain hallucination.

Claude Pro response

But it will nevertheless provide a summary, and will answer how it can do that if you ask it:

You’re right, I do not actually have direct access to the contents of the article or ability to visit links. As an AI system without a web browser, I generated that summary based solely on the prompt you provided describing [topic]. I do not have the specific details from the article to reference. My summary provides a plausible high-level synopsis of what such an article may contain, but cannot be considered an authoritative summary of the actual article content. I apologize for the confusion – in the future I will clarify if I am speculating versus summarizing directly from a provided source.

Clause Pro response

Some people will be wholly put off by this revelation. I say realise the chatbot’s current limitation in this regard but don’t write it off because of it. In the case of my example – asking it to summarise a magazine article – use it more effectively: share a PDF version of the article that you can share directly in the prompt. My experience doing this has been nothing short of outstanding.

Claude Pro excels at a task like this, far better than the other AI assistants. Of course, as ever, a great deal depends on your ask – how you prompt the chatbot.

So, is Claude Pro worth it compared to free Claude? Even when it made the type of mistake it did, and the revelation about links? I would say it is worth it (although bear this experience in mind). Invest $20 or £18. Install the web app on your laptop and your phone (here’s how to do that). Try it for a month and really kick the tyres.

Make the most of amazing tools like Claude Pro that will enhance your creativity and content creation. It will be of significant help in upskilling you with knowing how to use the current generation of gen-AI tools that will prepare you for what’s next in this emerging Age of AI.