Issue #1: Unfortunately, there are fundamental usability problems:

❯ In all other AI tools, you can have several sessions at the same time, i.e. ask a question, then ask another question in a parallel session and work independently in each session.
All other AI tools also allow you to retrieve each session and continue working EXACTLY where you left off.

» ChatPlayground AI does not support any of these functions:
• You can only work on one question at a time.
• If you open ChatPlayground AI in a new browser tab and ask a new question, you will lose the previous session.
• So you cannot reopen a ChatPlayground AI session and then have your originally selected AI models and all questions and answers again and can continue working there.

• If you want to find the history of a ChatPlayground AI session again, you have to search in the history of all AI models and hope that you can find it again.

❯ If you want to work with 5 AI models, ChatPlayground AI forces you to use 6 models. This means that you are wasting credits because every query with every AI model costs credits.

❯ It's nice to be able to ask 6 AI models questions at the same time, but reading the answers is VERY tedious when you can only see about 15% of each answer, even on a very large screen, and you have to keep scrolling: so it's not useful / efficient to compare answers.

» Every AI model box lacks a button that allows you to zoom the answer into a new browser window. With individual windows, you can position them in a way that is useful for reading and comparing.

❯ The vast majority of users want ONE answer in the end: they ask several AI models and work their way through the answers and follow-up questions until they have found / worked out a suitable answer. Unfortunately, ChatPlayground AI does not support this typical process. This should be possible:
» We should be able to temporarily enable / disable each AI model so that it no longer responds to follow-up questions.
» We should also be able to hide / show each AI model, so that if one Model is no longer of interest, it does not take up any space on the screen, but you can also show it again if you want to look something up.

❯ If you receive answers from e.g. 4 AI models, the time comes very quickly when you may only want to continue working with two AI models and ask them different follow-up questions.
However, ChatPlayground AI does not allow:
1. to deactivate individual AI models for follow-up questions, because you still want to keep their answers, but you don't want to ask them any more questions.
2. optionally ask follow-up questions directly to only one AI model: you cannot open a question text box for each AI model.

Issue #2: Security
I'm the IT manager for a company with several hundred employees - and one thing is COMPLETELY clear: We would have to strictly and harshly prohibit any employee from using ChatPlayground AI for work.
The explanation is very simple: ChatPlayground AI was looking for the cheapest way to give customers access to the portal by using 'magic links' and ChatPlayground AI ignores modern, minimal security requirements for cloud services.

This is the background:

The industry realised long ago that email inboxes are not secure: millions of hacked inbox credentials can be found on the Dark Web.
How has the industry responded? They have introduced MFA, multiple authentication, which means that the user has to prove that he has an ADDITIONAL security factor.

How have providers like ChatPlayground AI responded to the security problem? By REMOVING security factors: You must use a "magic" link to login.
ChatPlayground AI no longer even allows you to protect your account with a username and password.

The result?
ANYONE who gains access to your mailbox get access to ChatPlayground AI with the "magic link".

The problem?
❯ Hackers gain SILENT access to the user portal: The user is not aware that hackers can see and read the user data.
❯ For hackers, this is perfect for social engineering (what are the user's hobbies, interests, activities, etc.).
❯ For hackers, it is perfect for industrial espionage if the user works with documents, sensitive texts, source code, etc.
❯ The user cannot even explicitly protect the access to ChatPlayground AI: If someone has access to the mailbox, they have access to the ChatPlayground AI.
❯ So, the user cannot change a ChatPlayground AI password every 4 months, because, for example, this data is much more sensitive than the emails and needs additional security.

If ChatPlayground AI has a username and password but refuses to offer MFA to protect its customers, then a hacker could indeed hack the mailbox and then reset the ChatPlayground AI password.
But then, the user will realise that his account has been hacked because his password no longer works.

But ChatPlayground AI does not even allow customers to recognise when strangers are misusing their account.