Can you trust Gen. AI in decision making?

Artificial Intelligence

Ali Vakilzadeh Lead Project management officer| GTT Holding Tehran, Iran (Islamic Republic of)

Gen AI has demonstrated unfathomable features that have already dazzled everyone's minds. But it has also shown severe fluctuations in its way of conclusions, for example giving long, convincing descriptions for its faulty suggestions. How can a project manager trust such technology for decision-making purposes? If you can somehow reduce its failure rate, will you employ it to help with your decisions?

Posted: Dec 15, 2023 12:06 PM

Sort By:

< 1 2 >

Sergio Luis Conte Helping to create solutions for everyone| Worldwide based Organizations Buenos Aires, Argentina

Dec 22, 2023 1:49 PM

Replying to Ali Vakilzadeh

...

I agree that current AI models heavily depend on input and training data to generate outputs. however, several "decision builder" algorithms may influence an organization's behavior without anyone noticing the situation. For example, the ATS uses AI to summarize and rank applicant CVs, therefore a lot of applicants lose their chance of being seen, only because their CV wasn't "AI optimized". This means that an organization is not hiring the best candidate, but only the one selected by AI. Many other examples may exist, that negatively impact organ behavior without leaving any footprint.
In such cases, we are ignoring the low quality of output, just because we have little time to qualify the output.

You hit the nail: they rank the CVs and they notice about that. The decision is still on hands of human beings.. In Why? Because the AI can not understand what the word "best" and others means except you trained the AI with that and when you trained AI with that then you are including the organizational definition about the word "best". I created lot of algorithms on this matter because they are the most simple algorithm you can create and in fact people with knowledge about programing and some statistics techniques can create it just in few hours. This type of things are critical to understand and, in my humble opinion, organizations understand how the AI works when they try to use it. We need to understand that we, as human beings, are in control of AI. The problem, and please do not think I am writting this because of your comment, AI has become a buzzword like other things in the market: agile, big data, etc, etc.

Posted: Dec 25, 2023 7:29 AM (Updated by author: Dec 25, 2023 7:33 AM)

George Freeman Thought Leader | Author | Architect| Florida, United States

Hi Ali,

Regarding the portion of your question asking, “How can a project manager trust such technology for decision-making purposes.”

Consider the following:

1. Transparency » Leads to facts (and enables accountability).
2. Facts » Lead to the discovery of truth.
3. Truth » Authorizes trust.

If these dependencies are correct, we should ask ourselves, is there true “Transparency” in generative AI processes, frameworks, and metadata (control and learned)? If there is NOT, how can we lend our trust to these types of technologies?

Posted: Dec 25, 2023 5:35 PM (Updated by author: Dec 25, 2023 5:36 PM)

Ricardo Sastre Martin Principal Consulting Project Management| Microsoft Madrid, Spain

Trust will depend on the data sources used for providing the answers, and some tools currently does not provide any reference.

Posted: Dec 26, 2023 9:21 AM

< 1 2 >

Please login or join to reply

Can you trust Gen. AI in decision making?

Sponsors

Vendor Events

Guessing is not a strategy: How to build decision velocity with AI and real-time data

Newsletters