Issue link: https://maltatoday.uberflip.com/i/1531177
10 OPINION maltatoday | WEDNESDAY • 8 JANUARY 2025 ChatGPT has captured everyone's attention YOU can't turn a corner with- out hearing about artificial in- telligence (AI), particularly gen- erative AI tools like ChatGPT, Dall-E and many others. Wheth- er creating content, writing code or answering homework ques- tions, ChatGPT has captured everyone's attention. But, if you think that's the peak of AI's po- tential, think again. This year has brought a wave of new AI technologies quietly shaping the future. According to the Consul- tancy Firm Gartner, we're in the middle of a massive shift. While generative AI is rid- ing high, it's not the only thing making waves. Other AI tech- nologies are emerging, ready to make an even more significant impact on industries, business- es, and our everyday lives. If you're wondering why there is so much excitement around AI right now, you need to re- alise that AI is more than just a tech trend – it's changing the way we work, learn, and inter- act. From improving customer service with chatbots to diag- nosing diseases in healthcare, AI makes life easier, faster, and in some cases, safer. Generative AI, like ChatGPT, has been great at creating con- tent quickly, but the excitement around it also raises questions about its limits. For example, while it can write essays, answer questions, or even make jokes, it sometimes struggles with more complex reasoning tasks. This leads to what Gartner calls the "Peak of Inflated Ex- pectations" – a stage where the hype is massive, but real-world results haven't entirely caught up. We're seeing that right now with generative AI. It's fascinat- ing, but it's not the whole story. Beyond the usual hype, there's much more than meets the eye. Take something like synthetic data. AI needs loads of informa- tion to learn and improve, but sometimes, this data can't be shared because it's private. For example, hospitals have strict rules about sharing pa- tient data, so how can AI learn to spot diseases without seeing real medical records? Synthet- ic data is the answer. It creates fake but realistic data that AI can use to learn without break- ing any privacy laws. Think of it like making a train- ing video for a worker without using real customer informa- tion. This allows AI to learn and improve without the risks of handling sensitive data. Another compelling devel- opment is the rise of founda- tional models. These robust systems, far more powerful than ChatGPT, are being de- ployed across various industries to make critical decisions and solve significant problems. For instance, banks are lev- eraging these models to de- tect fraudulent transactions in real-time, while scientists are using them to predict weather patterns with unprecedented accuracy. Unlike generative AI, which is primarily focused on text and image generation, foundational models are empowering busi- nesses to make smarter deci- sions that can save time, money, and even lives. On the other hand, we have a growing army of AI tools designed to run in the back- ground, silently improving how companies and even govern- ments operate. These are called operational AI systems. While you might not see them at work, they help manage everything from supply chains to customer service depart- ments, making sure things run smoothly. Imagine a supermar- ket using AI to predict which products will sell out first based on the weather forecast. It can order more of those products automatically, so the shelves are always stocked with- out human intervention. This is the future of AI – working quietly behind the scenes to im- prove our daily experiences. One of the most intriguing concepts in the Hype Cycle is AI TRiSM, which stands for AI Trust, Risk, and Security Man- agement. As AI assumes more responsi- bilities, from driving our cars to analyzing our personal data, the need to ensure the safety and reliability of these systems be- comes increasingly critical. AI TRiSM is about instilling trust in AI systems, ensuring that their decisions are fair, trans- parent, and secure. For instance, in the context of hiring decisions, AI TRiSM guarantees that AI doesn't dis- criminate against any particular group. It's about ensuring that as AI takes on more pivotal roles, it does so in a manner that bene- fits everyone. We're also seeing exciting de- velopments in the education space with tools like AI makers and teaching kits. These are de- signed to help non-experts – in- cluding students and everyday people – understand how AI works. Imagine being able to build your own simple AI at school, learning how these systems think and make decisions. This kind of hands-on education can prepare the next generation to not only use AI but also shape its future development. It's no longer just about sci- entists in labs; AI is becom- ing something we can all get involved in, regardless of our background. So, what's next? While the cur- rent AI craze focuses on flashy technologies like ChatGPT, the The next wave of AI Prof. Alexiei Dingli is Professor of Artificial Intelligence Alexiei Dingli