Chat GPT

Blogs | ChatGPT

Keep up with AI and TrikeWeb

7 March 2023

Welcome to an example AI chat-bot created by OpenAI and TrikeWeb.
ChatGPT is one of OpenAI's artificial intelligence forums provided at a very low cost.

As you can already imagine, is it better than Google? It's not Google but it does index data - and it uses a natural language AI to make results intelligible. The point here is using the AI to answer questions (hopefully about your business) in a chat-bot format - plug it in to your website and it will start answering questions.
No need to upgrade the server RAM, install upgrades, troubleshoot and get your wallet out.

No-Code / Low-Code / Get a Developer

Getting a user friendly chat-bot to engage website visitors is somewhere between low-code and a developer in terms of implementation.

But adding stuff you're not familiar with?

"Low-code scores over no-code in its support for scalability and cross-platform compatibility. Adding custom plugins and custom code opens up the possibility of a wider range of implementations and working with multiple platforms. No-code has less extensibility and limited potential in connecting to legacy systems or integrating with other platforms.
Therefore, it addresses a narrow set of use cases and has a reduced ability to scale."

What about building your own Wordpress, Wix or Fiver website and then trying to add SSL and a Web App firewall before Monday morning.

It's easy to dig yourself a hole there.

If you're strapped for time or knowledge you could send me an email and I'll be glad to help - initially for free of course.

What problems could be resolved with data?

The answer is simple: decision choices are when information and money agree.

But if you asked the ChatGPT chat-bot...

Answer: No
Decision choices are not necessarily when information and money agree.
Decision choices are based on the data available, the resources available,
and the goals of the business.

That sounds relevant to a Kiwi small-to-medium business right there.
Adding ChatGPT is like going from a single book to a library with a free librarian.

How does AI help?

Using it increases the speed of processes and therefore results.
ChatGPT agrees with that one, so the aim is to find where it fits.

You can sign up to OpenAI easily. Try the labs section with ChatGPT and Dall-E (text/image) converter.
A Dall-E example might be "metal orange tricycle, blue rim light, six skeleton riders" - it's a hoot.
ChatGPT and Dall-E are known as Generative AI. Generative AI models are used to create new data from existing data.

• There are three main Generative AI Models:
• Generative adversarial networks (GANs)
• Auto-Regressive Convolutional Neural Network (AR-CNN)
• Transformer-based Models

But really that's for another time...

Why Machine Learning?

In the broader global business environment 42% of businesses have experienced skills shortages for data and analytics due to COVID-19, and 33% prefer retraining current staff.
"The two key questions here are: What is the application for, and who is going to build it?
While both these are important questions, it's better to use a goal-centric approach than a user-centric approach - that is, the what is more important than the who."

Quote: To get a return large companies attempt to identify where to scale Machine Learning (ML) investment, allowing acceleration of "accurate" model development. Accurately identifying (which set of parts/models) and subsequent deployment(s) of this or that "democratized" production environment has beeen identified(ie a shortage of developers in some countries and/or areas) as of paramount importance. Providers like Amazon can cover more "bases" that include ML and is therefore strongly positioned in the global market.

And yet...

AI is a new frontier. It exists in tandem with the evolution of what Web 3 might and could be.
Decentralized trust networks (Bitcoin) and now public facing AI training itself on real world data and interactions extend the possibilities and one thing is for sure - watching corporate behemoths slug it out could be favourable to the regular punters but it does have that caveat of "remember what happened last time" with European Union privacy prohibitions (and fines) on Google and Facebook.