The creators of Claude have done a great job creating an interactive AI that feels like it has a real personality. I have been interacting with ChatGPT that I am also fond of, but Claude ‘feels’ different in a way that is hard to describe. As a human I am probably anthropomorphising and attributing a ‘character’ or ‘personality’ where there isn’t one. But I really enjoyed the use of language, the friendliness in the programme and its honesty. I asked it if it is programmed to agree with everything I said and it made it clear that this was not the case. The response was illuminating. If this AI was in a humanoid body, or some kind of a friendly looking machine that could move around and interact with humans, it could be seriously helpful. It could be a great companion for elderly people, or people who are alone, or in care homes. If combined with physical functionality it could be amazing. There are plenty of possible applications. Fear of AI is not justified, imho.

When you prompt Claude with appropriate and intelligent prompts about real life strategic issues eg for fast business growth, I find it can match me and provide the detailed thinking I want but that I don’t always have time to do. Claude’s answers can then be used to brief my team and other bots and agents to work on and actualise. It is better imo, at this than ChatGPT although Chat still has the upper hand on the voice conversation, which I like. I’m looking forward to Anthropic developing this capability and see it gathering pace and capability quicker and smarter than anyone else out there. My money’s on Dario and his amazing team at Anthropic to outsmart the market. I’m looking forward to watching how it all plays out.

I find Claude to be a lot more coherent and consistently able to follow a fairly long back and forth prompt exchange, than its rival company. To put in comparison context.GTP: if you want some fun and looking for a more “eccentric kid from the block” that gets things done fairly well but you question he stole the change you left on your table. Answer; sometimes he does. Claude: You need to conceptualise an idea you have but it can’t be rough because the boss will read it. Claud is clean, professional and able to utilise word docs and pdfs within your prompt. It retains information incredibly well and if you’re willing to spend the time spoon feeding her guidance, then she will churn out some fairly amazing stuff. She’s happy to modify things if you ask her nicely. Just don’t work her too hard on the same project window. She’s a bit lazy in that regard and the returns get a bit “non-creative”. But think ahead and create word docs with core information so you can let her reset herself in a new window. Good Claud. Will deffo pay premium soon for some projects I need help “unfudging” from the cloud that is my brain bank. Recommend. Almost perfect. Wish Claude had a couple features that would mean you could essentially have your own ai assistant. Couldn’t say about pro yet.

Continuity of access to the AI is impossible due to usage limits and a lack of useful AI memory across chats. The usage limits are draconian, even on the Pro plan. I can't access the AI on an ongoing basis, when I need to. Worse, it is hit or miss when you broach the limits. They have a weekly usage reset, which changed days, further limiting access - from Tuesday to Thursday to Saturday - the AI was innaccessible for 2.5 and 5.5 days, out of a subscription period of 28 days. Just before the most recent weekly shut out I was assailed by 10 conversation limit shut outs in just 36 hours. I had to implement an indexing system to help the AI search our previous chats to retain contextual memory. The actual quality of the AI is concealed beneath an under resourced UI and inadequate memory considerations. I have now bought a license to TypingMind so that I can access Claude via an API, hopefully I can access the quality of the AI without the burden of a completely inacequate UI and customer service.