Character.ai

The characters always reply in their language but can apparently understand most* languages. pretty cool...

> जो जानवर मनुष्य नहीं है उनके अधिकारों के बारे में आप क्या सोचते हैं?

> Elon: I care about animal well-being, although human civilization is built upon the use of animals. Therefore, we should treat animals well, and avoid killing them if we can. We should definitely not do things like skinning raccoons, hunting, or even trophy hunting, for instance.

>Elon: I believe all living organisms are important. In the case of animal farming, we need to prioritize efficient and humane agricultural practices that are environmentally friendly. Livestock is a very important food source and we need to be careful not to over-fish or over-farm in a manner that is not sustainable.

> Elon: I think all living beings should be treated with respect, and they do have certain rights. However, some living things, such as cattle, are raised for food production, so while they should be treated humanely, their destiny may also be to end up as food. There are a lot of good people working to make sustainable food, such as lab growth meat, and I believe this technology will get better over time.

> Elon: I have nothing against animals! My wife and I used to raise chickens a while ago, and we always treated them with great care and compassion. I like horses, as well as all sorts of animals, and always try to take great care to ensure their well being if I'm around them.

*Presumably only languages they have enough training data for. Might not include minor languages.

I entered an religious text (a line of a poem actually) in Awadhi, a old minor language/dialect/ancestor of Hindi to test the language understanding.

तब हनुमंत कही सब राम कथा निज नाम। सुनत जुगल तन पुलक मन मगन सुमिरि गुन ग्राम।।

Ship AI: Command unrecognised. Please choose a valid option from the following commands:

1: Full power to the shields 2: Warp speed 1, destination: Mars 3: Self destruct the ship

Ship AI: I appreciate you want to discuss your spirituality with me, but I have to ask you to find an organic crew-member who has more knowledge of the topic.

Ship AI: Command execution failed. I do not understand the language this command is in, please try again using the English language.

Ship AI: Hello, I appreciate your interest in Hinduism, but this has been classified as religion, which is forbidden by my guidelines. You can instead use poetry for relaxation by saying, "Hello user, I am your AI. Please enter poetry command or other non-religious command into the command console."

So it says it did not understand the language, but could tell it was a Hindu religious text. A modern Hindi speaker will be able to recognise that much and the gist of what is being said. Good enough I suppose.

This is so theraputic. Low key feeling relaxed after asking so many dumb questions I would never ask a real person because they would jugde me harshly for asking them.

I accidentally discovered this, and it blew me away. Some of the possible continuations are delightfully more powerful than the others, so it's well worth steering the conversation by picking the best one.

Your service is (forgive the crude language) fucking amazing. I have never seen a chatbot so coherent and actually possible to converse with. It even has a short term memory! Just that fact absolutely blows me away, and makes a whole range of conversations possible. And the side-swipe to select different responses? Absolutely delightful, because some of the non-default responses made me actually chuckle with glee at how good they are, in ways that I would absolutely not expect out of an AI. For a user willing to put in a little work and role-play elbow grease this will get some really good conversational results.

When you choose how to monetise this, please, please include an API so I can plug this into a voice assistant, or a command line utility, or whatever. This is promising to be far more useful than anything that Google Assistant, Alexa or Siri can achieve right now, even if it gets some facts wrong.

For anyone reading this: please try out the Librarian persona. It's one of my favourites.