The Impact of Large Language Models

Revolutionizing Industry. Learn where you can use Large Language Models

A human face made of interconnected dots, representing the connections of the deep neural networks that make Large Language Models

Large Language Models (LLMs) are THE hot topic of the year. If the name Large Language Models sounds unfamiliar to you, I’m pretty sure you’ve heard of ChatGPT, OpenAI, and Bard. People who don’t know how to code have gained access to a tool that allows them to build Proof of Concepts for ideas they’ve been meaning to test for years but haven’t had the capacity to do so. 

In this blog post, you will learn about typical use cases of LLMs in the industry – and what you need to be careful about when using them.

NOTE: A human entirely wrote this blogpost – yes, humans writing content is still a thing! If you like this human’s ideas, book a meeting!

Do you want to further discuss this idea?

Book a meeting with Paulo Maia

Meet Paulo Learn More

Copywriting Assistant with Large Language Models

Using Large Language Models to speed up writing blog posts, sales pitches, and ads is a widespread use case where these models excel. You can feed the tone of your content or specific keywords you want to mention in the generated text – and you’ll be left with a canvas that you can edit to your liking, saving you hours of work.

Risks:

  • Lack of most up-to-date data: LLMs are trained on data until a specific time so they won’t be aware of the most recent trends. As a copywriter, you should be sensitive to this and should iterate.  Note that you can include this data as context in your prompt to mitigate this. 
  • Distinct style: LLMs have a typical way of writing: to improve the style, you can feed the model previous content written by you to make it more similar to your writing style.

Learning resources:

GDPR-Compliant Data Generation with Large Language Models

In a world of increasing regulations in AI, generative technologies can be used to create data for faster proof of concepts, avoiding training models from scratch (model cold start). 

Picture this: you are managing the customer service department and can train a model for text classification without using any personal user data, by generating various categories of complaints with LLMs! 

Risks:

  • Generated data is biased towards what is commonly seen on the internet: if you try to generate medical imaging data, you won’t have something very accurate and diverse! 

Resources:

Actionable Knowledge Base using Large Language Models

Being able to answer questions based on a knowledge base saves tons of time when onboarding a team member or when you’re trying to search for information in multiple unstructured documents. 

Risks:

  • LLMs might hallucinate information in some cases, when the answer is not clearly defined. 

Resources:

Coding Assistant

If you’re a coder – most likely you’ve used LLMs to get you unstuck on a piece of code or to learn a new language. This is a good complement to checking Stack Overflow for someone that might have had the same issue as you – but in this case, you won’t be left without a reply!

Be on the lookout for the code you submit here – a recent article has stated that Samsung workers accidentally leaked trade secrets via ChatGPT!

Risks:

  • Non-compiling code: There’s no guarantee that the generated code will 
  • Knowledge obsolescence: Junior coders end up being too dependent on these tools (but remember the last time you did a difficult mathematical operation by hand?) 

Resources:

Conclusion

Large Language Models are here to stay. Every week, there’s improvements and new tools built on top of LLMs – but we are just now starting to understand how to build products based on it.

Besides the obvious risks – we can’t forget about the extra dependency you’re creating for your business on OpenAI’s technologies. Any legal change (see Italy’s ban on ChatGPT) can completely break your business if you create a 100% dependency. Diversification of the used technologies is always good for mitigating risk. You should consider open-source alternatives for LLMs, if needed!

If you’re interested in using this new technology in your business, contact us!

Do you want to further discuss this idea?

Book a meeting with Paulo Maia

Meet Paulo Learn More

Like this story?

Subscribe to Our Newsletter

Special offers, latest news and quality content in your inbox once per month.

Signup single post

Consent(Required)
This field is for validation purposes and should be left unchanged.

Recommended Articles

Article
How Often Should You Retrain Machine Learning Models?

A common question in the domain of AI and machine learning is: how often should you retrain machine learning models? The answer isn’t as straightforward as you might think. It’s not a one-size-fits-all solution, but rather a process that requires careful consideration and strategic planning. In this article, we’ll explore three strategies for deciding when […]

Read More
Article
Reject Option: Your AI Model Has the Right to Remain Silent

When it comes to AI models, we often expect them always to provide an answer. But what if we could trust them more when they choose to remain silent? This concept, known as the ‘Reject option‘, allows AI models to abstain from answering when they are not confident, opening up many applications in your business […]

Read More
Kelwin Fernandes at the leacture the 3 tribes of AI. UPdate2024
Article
Empowering Future Leaders with AI: NILG.AI at UPdate 2024

At NILG.AI, we were recently proud sponsors at UPdate 2024. We lead both a workshop and a lecture by our CEO, Kelwin Fernandes, who has over 15 years of experience in AI. Our goal? To empower future leaders and professionals to navigate the complex yet rewarding terrain of AI with confidence and safety. UPdate 2024, […]

Read More