The Complete Guide Using ChatGPT And Other LLMs, to Boost Your Business (PART 2 OF 2)

Covering the OPPORTUNITIES, THREATS and an LLM ACTION PLAN.

Complete guide to LLMs - Part 2. Art by Midjourney.

Hope you had a good Easter break if you celebrate it.

This is PART 2 of a 2-part post covering a deeper (than normal) SWOT analysis on Large Language Models (LLMs).

You can read PART 1 by clicking here.

In this week’s post, I’m covering the Opportunities and Threats of LLMs to your business.

At the end of the post, I’ll conclude with the Future of LLMs and an Action Plan for you and your business to help survive the Age of AI.

So buckle up, we’re going deep.

---

Last week already seems like an age ago in AI terms, so let’s have a very quick recap of Part-1, you can read the whole article here if you missed it.

I identified 7 key Strengths of LLMs, as summarised by the diagram below.

From the well-known “text generation” made famous by the likes of ChatGPT to “semantic search”, which requires additional technology like a vector database to store embeddings.

The 7 main strengths LLMs

In addition to the strengths offered by LLMs, they also have some weaknesses.

The 4 main weaknesses I identified that users commonly experience in LLMs can be summarised as follows:

1. Hallucinations

2. Math and logic flaws

3. Bias

4. Lack of contextual understanding

I won’t go over these again.

If you didn’t read the previous article or need a reminder, you can catch up here.

So what are the real-world opportunities, the “killer-app” use cases of LLMs in a real-world business setting?

And, what are the threats they pose to businesses and society over and above their already described weaknesses?

Let’s dive in.