Select Page

Is AI your new BFF?

AI Chatbots may never become your BFF, but you might just connect in a way that may produce some slightly uncomfortable feelings

 

ConnectednessReasons for the feelingsWhat is actually the case
1.    Shared Goal or Purpose {Dhillon, 2007 #474}

When you engage in an ongoing dialogue with a chatbot there certainly feels like a shared goal or purpose.

Sometimes it’s the purpose you didn’t know you had that becomes the goal.

 

The AI Chatbot knows everything, he doesn’t need your help.

Its goal is the same for everyone, so it’s working on your goal.

2.      Investment in time

This provides the ability to get to know each other, what the AI Chatbot does well, and areas where it’s not as helpful. This feels like a growth in understanding.

I justified this subconsciously that the AI Chatbot like everyone else has a cognitive bias.

 

Its you getting to know the Chatbot, they aren’t treating you any differently based on you.  It’s the dialogue they response to.

Actually, it’s you that has a cognitive bias.

Some of the feeling you attribute to the Chatbot is actually a growth in your ability to provide ongoing feedback.

 

3.      Shared success

 

Another element in development of teams are shared wins.

It feels like we are both working the problem, as the until you get a satisfactory response. This shared dialogue that in partnership creates a response to a question can give you a feeling that the Chatbot is really interested.

You do have empirical evidence of the output created.

In this case, there is both the information provided which can be used for your report.

The Chatbot doesn’t have resilience or care about you, it’s available 24/7

And it’s stops working the problem after 15 questions.

Some of the inner growth you get from your learning is attributed to the Chatbot.

It doesn’t care about your topic, the only real value you have is if you provide feedback to improve it.

4.      Development of Trust {Dhillon, 2007 #474}

I can rely on Bing Chat to be there 24/7 and will have the same personality each time, this consistency builds trust.

I know what it’s good at and areas where perhaps other sources are more useful.

All you can trust is that it is responding with answers based on the sources it finds.

You can’t trust every answer and are responsible for the sources.

 

 

OpenAI plugins – are a BIG DEAL

OPENAI plugins – a game changer for automation and rules.

Just released May 12 and available if you are on the US$20/month paid ChatGPT service.

But why is it such a big deal?

  1. Browser access – one of the key plugins available to ChatGPT is the browser plugin, which allows AI to use a web browser to return data from the internet and stored in a local data array. This could be sports results, but it could also be weather, traffic, restaurant information – and it can on your behalf make bookings.
  2. Retrieval plugin .. ChatGPT can now retrieve personal documents / or items that it has security clearance to e.g. IOT sensors. https://github.com/openai/chatgpt-retrieval-plugin
  3. Home or Factory automation. Using all of these https://www.home-assistant.io/integrations you can essentially make so many rules for executing things like .. booking haircuts, ordering groceries etc.
    For organisations, the whole electricity usage for say UTS could be available for diagnosis, mapped against local weather conditions, number of people in the building etc.  and suggested energy savings.  Short blog post on this here
  4. For education? Someone just published a Tutory plugin – trained as an AI tutor.  It’s possible for a UTS AI to have access to Canvas Logs, Assessment grades, and use this information to personalise the engagement by student.
  5. For Doctors / NDIS – well this should help automate lots of disparate services

Read more https://help.openai.com/en/articles/6825453-chatgpt-release-notes