Do you use ChatGPT or similar AI tools in place of a good old-fashioned web search?

Can I Trust What ChatGPT Is Telling Me?

ChatGPT is more accessible than ever before.

ChatGPT stating that bad information is not intentional

you might evenuse a Shortcut to access ChatGPT via Siri on your iPhoneor Apple Watch.

ChatGPT isnt the only game in town either.

So easy that you might prefer it over performing an old-fashioned web search.

ChatGPT-3.5 stating that it is unable to reveal its sources due to tech limitations

But chatbots are not necessarily bastions of the truth.

However, theyve yet to earn our implicit trust.

We know this because the internet itself isnt trustworthy.

Google Bard’s attempt to verify its claims

Theres plenty of evidence of this where ChatGPT gets something hilariously wrong.

Whats not so funny ishow confident chatbots can seemwhen theyre wide of the mark.

Sometimes, chatbots make mistakes becausethe input you have provided simply isnt good enough.

ChatGPT recipe for Rustic Brick-Baked Earthy Delight

ChatGPT-3.5 sidesteps the process of showing a source for its claims in favor of a more conversational format.

This feels more like talking to a person than interfacing with a search engine.

Being aware of the source of your information is vital if you want to trust what youre reading.

For example, the article youre reading right now is sourced from How-To Geek.

The user had asked ChatGPT a question and been given an answer they were unsure of.

We’ve chosen not to link to this thread here due to the sensitive nature of the topic.

This is a dangerous intersection of bad data and health advice.

This person was making decisions relating to their health that were predicated on bad data.

ChatGPT didnt cite a source that could have hinted to the individual whether it was trustworthy or not.

At some point, you were probably urged not to blindly trust everything that you read on the internet.

Never Google a health complaint if you dont want Dr Google to tell you the very worst news.

Be aware that ChatGPT and similar tools may be trained on examples of this untrustworthy information.

Chatbots that obscure sources make this very difficult to track.

In the example above, I asked ChatGPT to name the largest bird in the world.

The response is not wrong, but when pushed for a source the chatbot comes up short.

This is especially true when they wont even tell you what sources theyre using for their claims.

Perhapslet ChatGPT take control of whats for dinner tonightinstead?