2023 was the year of Generative AI and exploring its possibilities in business. AI made headlines multiple times, breaking records and being a leitmotiv of multiple discussions in different fields. Global Investment in AI grew even before the rise of ChatGPT which leads to conclusion, that the hype has just begun.

Here are our top 10 predictions for GenAI in the upcoming 2024.

1. THE HYPE WILL NOT END

Usually, when something is that popular, the audience will get bored and stop talking about it relatively fast. GenAI might take another route because the discussion about it goes on, as business owners wonder how they can use it. Different organizations are checking different ways of adopting GenAI in their business model and that particular show will go on.

2. COMPETITION IN AI WILL INCREASE

It’s expected that the “AI wars” are going to heat up to fulfill the demand for new tool, toys, and business solutions. Google has already announced a new product called Gemini Ultra. There is however some skepticism present on the subject, whether it’s going to be more effective than the current version of ChatGPT. That being said – OpenAI is also expected to launch a newer version of their tool – GPT-4.5.

3. MULTIMODAL AI MODELS

OpenAI’s product isn’t the only one that should be viewed as an advanced LLM. Meta’s Llama 2 and Mistral served the same purpose and were just as impressive. Multimodal AI models allow their users to mix different formats (audio, text, video…) and use them to generate new content. Such advanced algorithms can learn from those, predict, and produce high-quality content that can be used for a vast spectrum of purposes. 

Since multimodal AI models are already providing good results and have caught lots of people’s attention one can be sure that their rise to success has just begun and in 2024 there will be even more solutions available.

4. THE CHIEF AI OFFICER POSITION IS GOING TO BECOME MORE POPULAR

Artificial intelligence usage has become a priority for Fortune 500 companies and as this trend is expected to continue, there is a need for appointing a Chief AI Officer to manage different AI operations within an organization. A similar trend played out during the rise of cloud computing a decade ago when companies decided to hire  “Chief Cloud Officers” who were expected to help them navigate the strategic implications of the cloud.

5. CHATBOTS TURNING INTO NETWORK MANAGEMENT

A lot of companies have begun to use chatbots and other visual genAI-based assistants and it’s anticipated that the number of users is going to not only be doubled but also be used for other operational purposes, mainly network planning and optimization.

It’s speculated that a new type of AI factory infrastructure will be necessary to support the expected growth in users and in the number of areas being supported by GenAI.

6. THE RISE OF SMALL MODELS

2023 was a year of large language models which caught the interest of the public. In 2024 it’s expected to notice more small models, trained on limited sets of data, yet, being capable of producing high-quality content depending on whether or not the subject belongs to their scope. PHI-2 and Mistral 7B look particularly promising and their purpose is to be tailored for specific needs in specific areas of expertise.

7. AUTONOMOUS AGENTS

Autonomous Agents are basically a type of software created to achieve a specifically given objective. As the name suggests it’s supposed to work autonomously based on provided data. This type of software is developed through machine learning and through advanced algorithms they are capable of making decisions with no human intervention. They can be used to enhance customer experience, react whenever the customer needs support, and create savings for the company.

8. THE GAP BETWEEN OPEN MODELS AND PROPRIETARY MODELS WILL BECOME SMALLER

It’s widely expected that open models will become even more similar in quality to proprietary models. Even in 2023 open models Meta’s Llama 2 70B, Falcon 180B and Mistral AI’s Mixtral-8x7B became extremely popular and when it comes to quality of performance, they go hand in hand with proprietary GPT 3,5, Claude 2 or Jurassic-2. And the gap is going to be narrowed even more.

9. OFF-THE-SHELF SOLUTIONS IN AI

2024 might bring even bigger possibilities in terms of customizing APIs. Soon developers will be able to quickly customize specific products providing high-quality solutions for all businesses. For end users, it means more intuitive and adaptive applications, because of intelligent assistants and summarization tools that can access up-to-date business information.

10. LLMs FOR IMPROVED OPERATIONAL OUTCOME

LLMs are already able to analyze data – both historical and from sensors which will give them the possibility to predict potential failures and mishaps in maintenance. As a result, the risk of unexpected downtime is going to be reduced, and simultaneously the lifespan of critical infrastructure is expected to be prolonged. Optimization of inspections and maintenance work is guaranteed to improve how effectively those tasks are being carried out. As a final result, organizations can save a lot by using LLMs for that purpose.

Author: Cezary Dmowski, CTO at ITSG Global

Let’s arrange a call and see how we can help you using the latest GenAI.