So, what will be the biggest trends?

I was thinking of adding an intro - but the title kind-of speaks for itself. (just so you know, we’ve listed 6 trends identified by IBM, so get ready to scroll).

1st Trend. Expectation → Reality

Now that the dust is settling on the widespread recognition and public use of Gen-AI in 2023, we are starting to approach the topic Artificial Intelligence from a more realistic angle.

We’ve all seen those movies. AI Robots take over the world, come up with unbeatable war technology, kill all of humanity…

…and if we’re honest, when ChatGPT was openly released last year, we all thought it was a bit creepy.

Is AI going to replace academia, our jobs, our friends, us?

The short answer is: no.

(At least not yet 🙃).

One of the biggest learnings from 2024 so far, is that AI is not here to replace. It is here to enhance.

AI softwares are being integrated into various workflows to optimise operations and complement the work that we are already doing.

More and more use-cases are emerging, including in sales, customer service, finance, recruitment - all of them allowing organisations to be more efficient, leaner & improve the quality of their outputs.

2nd Trend. Multimodal AI 

You can look at Multimodal AI like a natural evolution of AI.

Just like when primates grew a thumb and started being able to grab stuff and build tools, Multimodal AI is the next level of Artificial Intelligence, significantly increasing its capabilities.

Multimodal AI can take multiple data layers as input.

Not too sure what this means?

Before: 1 Model = Can be trained by 1 type of data

Now: 1 Model = Can be trained by 2+ types of data

Multimodal AI offers more human-like perception and problem-solving capabilities, as it can process information in ways similar to how humans integrate multiple senses.

Ok - this significantly increases training and knowledge of AI Models - but how does it apply to you?

Well, Multimodal AI also improves the application of AI Models.

For example, OpenAI’s GPT-4 can process both text and images. When using it’s Agent (ChatGPT), you can move freely between language processing and image generation - so you can for example write a text prompt to generate an image answer.

But that’s not everything.

Multimodal AI has a wide range of applications, including:

• Computer vision: Enhancing object identification and context understanding

• Natural language processing: Improving sentiment analysis and language understanding

• Robotics: Enabling better interaction with real-world environments

• Healthcare: Processing patient data from multiple sources for improved treatment

• Autonomous vehicles: Combining sensor data for real-time decision-making

3rd Trend. Smaller Models are on the rise

Small Language Models (SLMs) are designed to respond to specific use-cases and perform specific tasks with a higher degree of efficiency and precision than their bigger counterparts (LLMs).

SLMs are perfect for specific use-cases, as they can reach a level of accuracy, speed, and efficiency that LLMs struggle to achieve.

For example, a domain specific Language Model tailored and trained on the legal industry will be able to navigate intricate details way more effectively and produce better and more precise outputs.

On top of that, they are more cost-effective, have potential for enhanced security, and can process requests a lot faster than larger models.

Are you still following?

4th Trend. Virtual Agents are now able to perform tasks

Nope, a Virtual Agent is not a chatbot.

It’s moreeeee.

As AI develops and incorporates new sources and datasets (including through multimodal AI), capabilities of AI agents move beyond the simple interaction or following of commands.

Virtual agents now directly relate to actual task automation.

This means they get stuff done for you (reservations, checklists, emails - you name it). And although they still might need human supervision to perform complex tasks, they are continuously getting better and better.

Honestly, you should get one. And we can help you find one.

5th Trend. Regulation is getting serious

Regulation is the topic no one wants to cover, but actually is one of the most crucial aspects of AI implementation- and is getting serious in 2024.

Amongst others, the EU Artificial Intelligence Act is set to become a global standard in the regulatory environment.

In short, this Act assigns AI to four risk categories.

1. Unacceptable risk = banned

2. High risk = subject to strict requirements

3. Limited risk = specific transparency obligations

4. Minimal / No risk = largely left unregulated

We will unpack this topic further in upcoming issues & also cover it extensively within the Forum itself.

Rest assured, we will make sure that all of our members & subscribers are made aware & stay up-to-date about set regulations and laws.

6th Trend. & Finally: Shadow AI is real

Just wanted to make you aware of this one.

Shadow AI is the (un-official), use of AI in the workplace, by employees.

Yes - without approval - like for example when sensitive client information is passed through ChatGPT to create a proposal.

Not. Good.

If you think your employees are not doing it, think again. Impatience = Need for quick solutions.

Without the right policies in place, this can lead to serious issues in security, privacy, compliance, etc.

The most straight-forward solution is the implementation of clear rules and guidelines around how open-source models are being used within your organisation.

It is also equally important to train your employees on the right use of AI and make them aware of any associated risks. 

There are multiple ways to navigate this, and implementing Corporate Policy, as well as adequate training and refreshers also form one of the bigger education angles and services we cover within the AIL Forum community.

Well, there you have it!

These are some of the biggest trends and things to look out for in 2024.

The Forum is a great space to ensure you are properly guided on your AI adoption journey. We are with you from the very start and help you navigate big topics such as the right solution, the right partners, & the right regulations and policies.

So, if you are not a member of the AI Leadership Forum yet, join the waitlist today and become part of the next cohort of applicants!