TRYEngineering - AI Adventures - 9

Large Language Models
Large language models, a type of generative AI you may be familiar
with is, focused on generating text to respond to a person's prompts.
This is what ChatGPT and Microsoft Copilot are! They were trained on a
large volume of text, much more than anyone has ever seen! They use
what they've learned, along with the " prompt, " and figure out the best
text to return. The volume of text they're trained on exemplifies big data,
a large amount of information that surpasses the capacity of humans to
sift through using conventional methods. An AI can process all that big
data and understand it, identifying patterns and information within it.
This is called data mining and is used to determine the best response
to any prompt!
AI Hallucination
AI tools, like ChatGPT and Copilot, are usually good at responding to
people's prompts, but sometimes they might give you information like it's
a fact, even if it's wrong or misleading. This phenomenon is known as AI
hallucination, and it typically occurs with large language models.
This is different from a deepfake. While AI hallucinations are unintentional
and not something the user wants the AI to do, deepfakes are deliberate
attempts to be fake or misleading.
There are a few reasons why this happens. There may not be sufficient
training data for the AI to correctly respond to your prompt. The AI might
be assimilating bias from the training data or misinterpreting the prompt,
resulting in a poor response.
While there are a few potential causes, it's important to keep in mind
that an AI tool doesn't think; instead, it utilizes its training data and input
to determine the response most likely to satisfy the user. From this
perspective, it becomes clear why AI hallucinations occur.
These hallucinations could have a greater impact as AI becomes more
prevalent in industries and fields like healthcare and finance. It's crucial
to comprehend the reasons behind these hallucinations in
order to reduce their likelihood. Even in personal use,
hallucinations can have an impact if an AI provides
someone with false information and they
believe it without verifying it. If you ask an
AI for facts or information, verify it with
another website or source to make
sure it's right.
Popular AI Tools
In 2024, we saw a rise in the use of AI
tools. Here are ten cool AI tools:
1. CapCut: Makes video editing easy
and fun
2. Character.ai: Creates stories and
chats with AI characters
3. ChatGPT: Helps answer questions
and chats about anything
4. Claude AI: Helps write stories and
work on creative projects
5. DeepL: Translates languages
very accurately
6. Google Gemini: Handles tasks
involving both text and images
7. Midjourney: Creates cool
AI-generated art
8. Microsoft Copilot: Helps answer
questions and chats about anything
9. Perplexity AI: Helps answer
questions and chats about anything
10. QuillBot: Helps rewrite and
improve your writing
These tools are making a big impact
in many areas, from creative arts
to productivity.
REFLECT: Which one interests
you the most?
http://www.Character.ai
TRYEngineering - AI Adventures

Table of Contents for the Digital Edition of TRYEngineering - AI Adventures
