AI Tools Takes (Transmission #330 Part I)

WRITTEN BY: SETH SIEGLER

As a follow up to the recent State of AI group discussion, here’s my promised follow-up: a list of the AI tools myself and the eXp team have been using recently, and a quick take on what the pros and cons are (from my experience):

ChatGPT (Enterprise version at eXp)

Good For: Closed ecosystem for GPTs for staff, enabling alignment and asynchronous collaboration (the main feature of Enterprise ChatGPT besides the data privacy aspect where OpenAI does not train on chats from Enterprise accounts. Otherwise it’s the exact same thing as ChatGPT Plus).

Bad For: Any sort of chat experience that you want to use in your own UI and also does not allow for MCP integration at this time. If you want to call an external API from a chat, you have to create a custom GPT and use the “actions” feature, which is not as handy as pre-built MCP integration. 

Notes: Beta customer for 2 years; used for internal operations and secure AI experimentation. We have 1,500+ staff members using it daily and over 1,000 custom GPT’s created and used to streamline and optimize work. 

Website: https://chatgpt.com/

Claude

Good For: Another opinion and perspective to ChatGPT’s output for important strategic brainstorming is to take output from ChatGPT and ask Claude of its opinion on it. It drastically reduces the “yes man” issue that all LLM’s have once they get to know you a bit. Also Claude has native MCP integrations so you can easily add external API calls and system integrations into your Claude chat experience. (ex: Chat with Claude and it will decide on its own to call and gather data from Hubspot, your email, update your calendar, etc)