Frank BacchusCIO LTS International / Chief Program Manager| LTS InternationalHolmdel, Nj, United States
Beware: Information generated by AI tools like ChatGPT, Gemini, and others may not be current. Double-check everything before publishing, especially facts and figures. A simple prompt asking "When was this data last updated?" can uncover surprising outdated information.
What is your thought on this?
frank Saving Changes...
Senior Projects Manager | Field & Marten AssociatesNew Westminster, British Columbia, Canada
Frank, good point. I think the last time ChatGPT was updated was January 2022 (Not 100% sure though). We have to always check the output from those AI tools, and never reply fully on the answer provided without verifying it. Saving Changes...
Always, always verify and validate. Sometimes Generative AI provides answers that are about as good as Karen on Twitter...
And if you have any doubt, challenge. You'll be surprised how often you can ''teach'' the AI... Saving Changes...
Dipalim KoushikProject Manager| Nielsen IQVadodara, Gujarat, India
Pascal, exactly. Generative AI is just like a baby whom you can mold in a way that you want it to work. They will provide you answers according to the way that you want. You just need to spend some efforts at the initial stage and later they will give the output according to your needs.
...
1 reply by Pascal Brunet
Feb 23, 2024 7:02 AM
Pascal Brunet
...
EXACTLY - If you have ever watched one of Ricardo Vargas webinars or AI masterclass (and if you haven't, I strongly recommend it!), he has this slide he presents quite often that resonated with me. When using AI, the first 10% of the work is done by you and influenced by the quality of your input and prompting. The next 80% is done by the generative AI. The last 10% is again done by you to verify and validate the results. This is also what PMI in their training refer to as the ''human in the middle'' part of using Gen AI.
Yes, exactly, and your point is really significant. Up until a predetermined deadline, ChatGPT and other AI models produce responses depending on the facts and patterns found in the data they were trained on.
Users should proceed with caution and independently confirm information, particularly when discussing urgent or quickly evolving topics. It's also a good idea to inquire about the most recent update or confirm the accuracy of the material, as you correctly pointed out. Saving Changes...
Pascal, exactly. Generative AI is just like a baby whom you can mold in a way that you want it to work. They will provide you answers according to the way that you want. You just need to spend some efforts at the initial stage and later they will give the output according to your needs.
EXACTLY - If you have ever watched one of Ricardo Vargas webinars or AI masterclass (and if you haven't, I strongly recommend it!), he has this slide he presents quite often that resonated with me. When using AI, the first 10% of the work is done by you and influenced by the quality of your input and prompting. The next 80% is done by the generative AI. The last 10% is again done by you to verify and validate the results. This is also what PMI in their training refer to as the ''human in the middle'' part of using Gen AI. Saving Changes...
Yes, you are right. They may be outdated. Saving Changes...
Sergio Luis ConteHelping to create solutions for everyone| Worldwide based OrganizationsBuenos Aires, Argentina
This is not new. When you use ChatGPT or similar, when you use a generative AI tool, you have to read carefully (or asking it as you stated) the last update on the information. Beyond that, few people remember that all they get from generative AI tools must be validated. Saving Changes...
Lavanya VijayaraghavanLearning Experience Designer| FreelanceBangalore, Karnataka, India
There's been a more recent update - In its own words, ChatGPT 4's last update was in April 2023 AND it's complemented by internet browsing for the latest information. I use ChatGPT Plus & Co-pilot to find information available on the internet and always ask for the link to the source. I have definitely seen hallucinations & links to totally unrelated sites. I try to verify facts with the original source as far as possible and of course evaluate the source for reliability (just because there's a source doesn't mean the information is legit - pretty much a no-brainer). Saving Changes...
Community Champion
