Generative AI: AI which talks, writes, draws, & more!
Most of us are aware of ChatGPT, which has been hogging the news for almost a year now. It belongs to a category of AI Models known as Generative AI. Let’s find out more about Generative AI in the following sections.
What is Generative AI?
AI, which generates content. Some examples below are based on the type of content.
How does it work?
Most of the generative AI’s work on a simple three step process as below.
What are other examples of Generative AI models for text data besides ChatGPT?
Here are some examples of open-source LLM (Large Language Models)
Vicuna, Alpaca, and Llama are Generative Pre-Trained Transformers (GPT) Large Language Models named after “South American” animals! Other models include T5, MPT, Falcon, Dolly, etc.
Generative AI in action: Stable Diffusion Model
Stable Diffusion is a generative AI model that can generate detailed images from text descriptions.
ChatGPT Simplifying Legal Jargon: An Example
Explain the following in easy to understand language “A certain tract of land situated in the Parish of Assumption, State of Louisiana, on the right descending bank of Bayou Lafourche, and forming part of the Trinity Plantation, and having a front on the public road of Bayou Lafourche of Two Hundred Eight and 70/100 (208.70’) feet with the same width in the rear, and running back from said public road, between parallel lines, the upper line of said lot measuring Nine Hundred and Eighty-two (982’) feet in length; the starting point of said lower line from Bayou Lafourche beginning a distance of Nine Hundred and Eighty-two (982’) feet from the lower line of Trinity Plantation, and which property is fully described on a plan of survey made by J. Martin Rhodes, C. E., on February 8, 1909, and annexed to and made part of an act of sale from Walter Guion, et als, to George Seth Guion on February 16, 1909; together with all buildings and improvements thereon and all rights, ways, privileges and servitudes thereto appertaining and belonging”
Response
In simple terms, it describes a specific piece of land located in Assumption Parish, Louisiana. The land is situated on the right side of Bayou Lafourche and is part of the Trinity Plantation. It has a frontage or width of 208.70 feet along the public road of Bayou Lafourche. The width is the same at the rear of the property. The land extends back from the public road in a rectangular shape, with parallel lines on either side. The upper line of the lot measures 982 feet in length.
The starting point of the lower line of the land is 982 feet away from the lower line of the Trinity Plantation. This information is based on a survey conducted by J. Martin Rhodes, a civil engineer, on February 8, 1909. The survey plan is attached to a document called the “act of sale” between Walter Guion and others, and George Seth Guion, which took place on February 16, 1909.
Additionally, the property includes any buildings, improvements, rights, paths, privileges, and legal agreements associated with it
Things which Generative AI is not good at:
MATH
PUZZLE
Limitations of Generative AI
- Cost: Generative AI can be expensive to implement. It requires specialized hardware and software, as well as skilled personnel to operate and maintain it.
- Time: Generative AI can take a long time to train and deploy. It’s not a “plug-and-play” solution, and it can take weeks or months to get it up and running.
- Data Quality: Generative AI relies on high-quality data to make accurate predictions. If the data is incomplete, inaccurate, or outdated, the results can be unreliable.
- Ethical Concerns: Generative AI can raise ethical concerns. For example, it can be used to make decisions that are biased or unfair.
Risks of Generative AI
Along with the immense benefit of Generative AI, there are many risks as well while adopting this technology which include, Deep Fakes, Fake News, Plagiarism, Ethical Concerns and Bias.
Applications of Generative AI
- Automating Writing Tasks: Helping users come up with better text content.
- Question Answering: Helping users find answers to questions with or without a knowledge base
- Better Chatbots: Improving conversational AI quality.
- Generating Codes: Writing good quality codes for all basic as well as complex tasks
- Text Summarization: Generating high-quality summaries of long documents etc.
The future for generative AI seems bright and plentiful. Keep a watch on the emerging trends in this space. To learn more about building hands-on skills in Generative AI, do check out our Masters program in AI (link here).