184. What AI will disrupt in 2024 and how to get ready

We are still at the beginning of the AI revolution, but in 2024 some of the world's oldest and richest industries are going to get disrupted by it. 

Listen to this episode to learn:

• The difference between Generative AI vs Large Language Models
• Why generative AI could get you sued
• Which industries generative AI and LLMs will disrupt in 2024 and why
• How to prepare for the LLM revolution in 2024 if you're an innovator, corporate leader or investor

This is part 1 of the 2024 tech predictions mini-series.

Join our January program & have your best year yet:

Design for Growth 

Resources mentioned in this episode:

Algolia: Large language models (LLMs) vs generative AI: what’s the difference?

Listen here on Apple

Listen here on Spotify

Join the Tech for Non-Techies membership and succeed in the Digital Age. 

Tech for Non-Techies clients 

• Reach senior leadership positions in Big Tech firms
• Lead digital transformation in established businesses
• Create tech businesses as non-technical founders
• Pivot into careers in venture capital

We love hearing from our readers and listeners. So if you have questions about the content or working with us, just get in touch on [email protected]

Say hi to Sophia on Twitter and follow her on LinkedIn.

Following us on YouTube, Facebook, Instagram and TikTok will make you smarter. 

Episode Transcript

Hello smart people!

Happy New Year! What do you think? Shall we make 2024 our best year yet? Let’s give it a go. Why not? You only live once, so we might as well give it our best shot.

So when we think about the year ahead, let’s remember that we do not have full control of our outcomes, but we do control our inputs.

And in general, when we have thoughtful inputs, we are more likely to have the results that we want. You know this with food, right? If you eat more greens and less sugar, you will feel and look better.

It’s the same with our brains and our actions: if we take the time to educate ourselves, and learn how to tell real innovation from tech hype, we are more likely to become leaders in the Digital Age.

This is why, my dear smart listener, in the next few episodes I am going to share my predictions for what will happen in tech in 2024, so you can be best prepared for this year.

In today’s episode, the first of my predictions mini-series, you’ll learn about the top tech trend that will dominate 2024, and you’ll see where this trend is dangerous and where it’s a huge opportunity. This way, you will be able to make the smartest decisions for your career, for your venture or for your investment strategy.

And obviously, this is not investment advice, so please don’t sue me. This is an educated opinion on what is on the horizon and how you can make the most of it.

And, in the name of controlling your inputs, have you registered for Design for Growth yet? This is the best way to start your 2024.

Design for Growth is a workshop that will take place online on 28 January and consists of two main parts.

First, you will learn design thinking principles, which will be covered by me and a very cool Lead Designer, and then, in the second part, we will use these principles to create a plan of action for your best year yet.

You can use the Design for Growth program to create a new product, to launch or grow a venture, or to transform your career. I have used design thinking for all three, and it works.

As well as this workshop, you’ll also get a class on how to get headhunted, so you can make opportunities come to you, which we will do in February. All of this is online, so you can grow your network, and learn new skills without leaving your house.

There’s also a bunch of bonus resources, which are all in the program description at techfornontechies.co/events and in the show notes

This is going to be a very small group, so everyone can get feedback and coaching, and make genuine new connections, so make sure to get your space before they’re gone.

And now, let’s start with prediction number 1 for 2024.

My number one prediction is that Large Language Models are going to continue rising in popularity, especially closed Large Language Models.

Let’s quickly define terms: a large language model uses natural language processing to understand and generate humanlike text-based content in response.

It’s AI for language, not images or music. ChatGPT 3 and Chat GPT4 are examples of Large Language Models.

DALL-E, which uses AI to create images is not a large language model, because images are not language. Got it?

So what is DALL-E? It is generative AI.

Generative AI is artificial intelligence focused on creating models that can make original content, including images, music, or text. They do this by processing lots of training data, and use machine-learning algorithms to understand patterns and then give you content based on your prompts.

LLMs, or Large Language Models, and generative AI are different parts of technology, but they get combined to produce content.

If this went over your head, don’t worry about it too much.

I pasted a good article with clear explanations of this in the show notes, but for the purposes of today, just remember that Large Language Models, are about language only, and generative AI is about language and all sorts of other things, which includes language and images and music.

So, my prediction to you today is specifically about Large Language Models, and the one you probably know best is Chat GPT.

ChatGPT is wonderful and it has already saved me lots of time, but LLMs do have their issues.

If you understand the issues, you will understand how Large Language Model trend will develop in the future.

One of the issues with LLMs is that they hallucinate - this is when the tool literally makes up information.

For example, I ChatGPTd myself, and got a bio of myself that sounded like it could be true, but it wasn’t. For example, it said that I work in a venture capital fund and have a book out with a famous publisher. This is close to the truth: I advise venture capitalists and I used to work in a private equity fund ages ago, and I have written extensively for elite publications like the Harvard Business Review and the Financial Times.

But I am not a venture capitalist, nor do I have a book out (yet). But the problem with generative AI hallucination is that it sounds true, it’s not like it said that I was a figure skater and mother of 5. So if you don’t know the actual facts, you could get into trouble when you only rely on ChatGPT. And there have been some notable examples of this, like the lawyer who ChatGPTd case history and presented it as true to a judge, and it was all made up.

Because in 2023, generative AI spread so quickly, the problems associated with it also spread quickly, hallucination being one of them.

Given that this is actually quite a hard technical issue to solve, I suspect that while LLMs will continue rising in popularity, we will need to learn to verify everything they tell us.

Another issue is the data itself, and that’s the data that goes into the model.

For example, if you use AI to make write a creative brief for a new advert, as in you write a script for a new ad, but the data that the AI uses is the original work of other creatives, those creatives might claim that their copyright is being violated and they could sue.

Will they sue the LLM company or will they sue the user? I don’t know, but in 2024 we are going to find out.

I’ve seen professional services firms and creative companies getting really worried about this issue, and for good reason.

Actually one of my clients is a lawyer with deep knowledge of technology and AI, who is now working on how to solve the issue of rights clearance in the digital world. So I am reporting to you from the frontlines of the AI legal battle, and I think we’re only seeing the beginning of this battle.

So, the result I’m seeing is that companies with deep pockets are investing in their own closed LLMs.

A closed Large Language Model is basically a system that is trained on the data set that you feed it, rather than the internet as a whole. For example, if you’re a huge old established law firm, you will have information on all the cases you’ve been involved in and you will know it is absolutely correct. You could use that, plus the data you have approved for your model.

I’ve seen law firms and advertising companies do this already, which makes sense - they have the data, the money and the incentive. The incentive is that their people are very expensive, and their growth is constrained by the human element - even corporate lawyers have to sleep sometimes.

So, if you can make their time more productive with a verified version of ChatGPT, you’re going to want to do it. This is why, this model will disrupt the professional services and creative industries first.

AI is not going to replace lawyers, just like it is not going to replace advertising creatives or doctors.

But, these professions are going to be seriously shaken up by AI, and the ones that flourish are those that make AI work for it, rather than ignoring it.

Since all of us have used lawyers and accountants at some point, I want you to know about this, both as a business leader and a customer.  Because you don’t want a lawyer who doesn’t ever use LLMs, because your legal costs will be higher than they need to be. But equally, you don’t want a lawyer who uses LLMs, without understanding its limitations, because you this could get into serious trouble.

More and more traditional professions, like lawyers and ad creatives will require close collaboration with technologists. This means that even those who least expect it will have to speak tech.

Thankfully, speaking tech is much easier than actually doing tech. Even after a one day workshop on Tech for Business Leaders, my clients have told me that they feel more confident in collaborating with technologists and speaking to their tech clients.

This is why, you organisation definitely needs one the top rated Tech for Non-Techies trainings.

AI won’t take jobs. It will be people who know how to work with AI who will take jobs.

Where there is change there is opportunity. If you are listening to this podcast, you’re obviously the kind of person who wants to focus on the opportunity of AI, and not pretend it isn’t happening.

And now I’m going to share some practical advice for steps you can take to make the most of the Large Language Model opportunity, as a user, as a corporate leader and as an investor.

•  Let’s start with users, that’s basically all of us. Firstly, you need to understand what Large Language Models are, know where the data that’s fed into them comes from and also be aware of hallucination. Know that you will need to check any facts that the model gives you, because they could be completely made up.
•  So, for 2024, I suggest you spend some time learning ChatGPT or another LLM.
•  For corporate leaders:
•  Have you thought about your AI policy? Do you have one? The thing is, someone in your company is already using ChatGPT, or another LLM, and that could mean copyright infringement. That could also mean that they might be falling for model hallucination, and not be aware of it. I am not saying don’t use these models, you definitely should, but you need to have a smart policy for how to do this and guardrails. These policies should be made up by people with experience, not some poor junior HR person. The LLM revolution is serious, so take is seriously.
•  Also, for the corporate leaders here, what’s your plan for closed LLMs? Services are popping up for specific industries, that are based on training data for your industry. I’ve already seen them for law and accounting, and expect more will pop up in 2024. Do you know how to get the right provider? Do you know how to ask them the right questions when you have several providers pitching to you? If not, this is a good place to start.
•  For investors: it does look like on the startup side, we are at or reaching an AI valuation bubble. I’ve seen tech hype before about other things, like crypto and the meta verse and NFTs. Now it looks like we are in the midst of AI mania. Unlike crypto and the metaverse, there are wide applications to generative AI and Large Language models. But, revenue is not the same as valuations. And what happens to valuations in a bubble? They go sky high and then they burst. So, when you’re looking at AI companies to invest in, be smart about it. look at the fundamentals. Find your inner Warren Buffet. If you don’t understand why someone would pay for a tool, don’t invest in it.

Tech for Non-Techies listeners are clever and discerning. You do not fall for hype. You use your own brain to make decisions.

And that’s why I love you.

That’s it for this week’s lesson. Tune in next week for my next prediction for 2024 and how to make the most of it for your career, your organisation or your investments.

And, if you are finding these episodes useful, tell me about it. Leave this podcast a rating and a review, as your New Year gift to me. I would really love to hear from you.

And, remember to sign up for Design for Growth, which is taking place in just a couple of weeks. It is honestly going to be a great way to start your year in the strongest way possible.

Remember what Benjamin Franklin said: an investment in education pays the best dividend.

So invest in your education today and join Design for Growth.

The link to sign up is in the show notes or just go to techfornontechies.co/events

Thank you very much for listening, my dear clever listener.

Have a wonderful day and I’ll be back in your delightful smart ears next week.

Ciao!