Incorrect or misleading labels/outputs are provided by the AI and Machine Learning Models

 

AI and ML products have proliferated as businesses use them to process and analyze immense volumes of data, drive better decision-making, generate recommendations and insights in real time, and create accurate forecasts and predictions.

 

Gemini is a large language model (LLM) developed by Google Artificial Intelligence. Gemini is built on a foundation of advanced machine learning techniques, including transformer architectures and deep learning. It has been trained on a massive dataset of text and code, allowing it to acquire a deep understanding of language.  

 

It is designed to be a versatile AI assistant capable of a wide range of tasks, including:  

§  Gemini can generate human-quality text, such as articles, essays, code, scripts, musical pieces, email, letters, etc.

§  It can translate text from one language to another accurately and naturally.

§  Gemini can write code in various programming languages, including Java, Python, JavaScript, and CSharp.

§  It can provide informative and comprehensive answers to a wide range of questions.

§  Gemini can summarize long texts into shorter, more concise versions.

§  It can generate creative content, such as poems, stories, and scripts.  

 

Overall, Gemini is a powerful and versatile AI assistant with the potential to revolutionize a wide range of industries.

Note: Google Gemini’s AI and Machine Learning Models are utilized for the below use cases.

 

Use Case 1:

The below two features are used as input to Google Gemini.

Ø  Flights list from New York to Mumbai

Ø  Air India Flights from New York to Mumbai cheap rates

 

v  For the first feature i.e., “Flights list from New York to Mumbai”, the below label is present by Gemini.

 

 

When searched the flights using the Google Flights link, two Air India Flights were displayed  and both were direct flights.

 

 

 

But when searched the Air India website, there were no direct flights from EWR – BOM and its details.

 

 

 

Google Gemini has provided incorrect or misleading information, as the content displayed by Gemini, that there were direct flights, which was partially correct, but it recommended to Google Flights Link, where there were direct flights from EWR – BOM, but the information in the Air India is completely different, and there were no direct flights.

 

v  The second feature i.e., “Air India Flights from New York to Mumbai cheap rates”, the below label is presented by Google Gemini.

 

 

Google Gemini content gave the information that direct flights may be more expensive than flights with layovers, but when queried the Air India flights, the opposite was true, such as the direct flight was the cheapest option, and the details goes here. The Google Gemini content was thus misleading.

 

Summary: The randomness, generalization, and pattern identifications techniques are still evolving, as it continues to develop, we can expect to see even more innovative and exciting applications of this technology.

 

 

 

Use Case 2:

 

The below feature is used as input to Google Gemini, as a fresh query.

Ø  Direct Air India Fligths list from John F Kennedy to Mumbai on 13 Oct

 

Google Gemini, gave the above label for the feature (Direct Flights), such as there are multiple flights from JFK to BOM and the number may vary. It is recommended to visit Air India Link for details. In the Air India Link provided by Google Gemini, there was only one direct flight from JFK to BOM, the details are as follows.

 

 

 

Here the content provided by the Google Gemini and the Air India link data are not matching, the information is misleading.

Hence, we can safely conclude that incorrect or misleading labels/outputs are provided by the AI and Machine Learning Models which we have to double check before arriving any conclusion by using AI for outcomes.