Hamilton Public Library's Chatbot helps Members find information about the library, community services, and resources. The Chatbot runs when the library is closed and Staff aren't available to answer your questions. The Chatbot has information about a limited selection of topics, so depending on what you're looking for, you may want to contact Staff when HPL is open. You can also leave a message and Staff will respond later.
Answers to Frequently Asked Questions
How do I know when I'm talking to a chatbot?
The Chatbot starts all conversations by telling you it is a bot. We don't want to trick anyone. If that Chatbot message doesn't appear, you're communicating with HPL Staff.
When does the chatbot run?
The Chatbot runs when the library is closed.
How does the chatbot work?
The Chatbot uses a probability-based system known as discriminative artificial intelligence. It has a limited set of answers to questions. When you ask it a question, the Chatbot tries to guess which answer is most likely correct by comparing your question to previously asked questions.
Can the chatbot make up its own answers to questions?
No. The Chatbot can only pick an answer from the set of answers HPL Staff has written for it.
Can HPL staff see what I say to the chatbot?
Yes. HPL Staff can see your conversations with the Chatbot. We use these conversations to train it to give better answers in the future.
If my conversations are being used to train the chatbot, what is being done to protect my privacy?
Freshworks provides the artificial intelligence (AI) model that powers our Chatbot, which is unique to the Hamilton Public Library. Freshworks has very strong rules and systems in place to prevent other customers — or its employees — from accessing HPL's AI model.
When HPL Staff train the Chatbot, it only changes HPL's AI model. Freshworks does not use HPL’s training data to change its AI model for other customers. If you want to know more about the security measures Freshworks uses to keep your data safe, you can read more about it at trust.freshworks.com.
Why don't you use a more powerful AI model, like some form of generative AI, for the chatbot?
Generative AI chatbots can be very powerful and can give very useful information, but sometimes, they say things that go against HPL's values: they might say something hateful or provide you with information that might not be true without being transparent about the source. HPL wants to be accountable for what the Chatbot says to you. HPL has chosen this AI model because it is the best tool to further HPL's mission: Freedom to Belong and Discover.