What exactly is artificial intelligence (AI)?

Artificial Intelligence (AI) has become an integral part of our society. Experts believe that AI has the potential to revolutionize many industries, which is why investors are investing huge amounts of money in it. But what exactly is Artificial Intelligence? Let's dive into it briefly.

Where does Artificial Intelligence come from?

In 1956, John McCarthy, an emeritus professor of computer science at Stanford, coined the term “Artificial Intelligence” at the Dartmouth Conference. At this conference, McCarthy suggested, “Research should continue on the assumption that every aspect of learning or any other property of intelligence can in principle be described so precisely that a machine can mimic it.”

What exactly is Artificial Intelligence (AI)?

The concept of artificial intelligence, often abbreviated to AI, is the subject of much debate and has no clear definition. Depending on who you ask, you'll hear different interpretations and theories. However, it is generally assumed that AI refers to computer-controlled systems that can take over tasks that previously could only be performed by humans. This involves taking over cognitive functions or human actions, so that computers are able to simulate both the human thought process and motor skills.

This includes a wide range of applications, ranging from systems that can recognize images on video images, speech recognition systems and systems that can make predictions based on extensive data sets, to advanced robots that can perform complex surgical procedures. It's essential to understand that AI isn't one specific thing, but a collection of technologies.

At The AI Group, we often use the metaphor of a large LEGO box. Each LEGO brick represents a specific technology or function. These “bricks” can be used independently, but can also be combined to build more complex and sophisticated systems. Consider, for example, an advanced smart doorbell that allows you to ask in plain human language what time the mailman was at the door. The strength of AI lies in its ability to combine these technologies to create solutions that were previously unthinkable.

What are the four phases of AI?

The developments surrounding AI can be defined in 4 different phases, all of which have had important breakthroughs:

Statistics with muscle balls (1960s)

‍In the 1960s, when computers entered the academic and research world, AI was primarily seen as “statistics with muscles”. This era was defined by simple algorithms and rule-based systems. This first generation of AI systems tried to simulate human intelligence through direct instructions and decision trees. Although the possibilities of these systems were limited, they did lay the foundation for future developments in the AI sector.

Machine Learning (around 2000)

‍At the turn of the century, Machine Learning (ML) emerged as the dominant approach for AI. Unlike the previous phase, where rules were set manually, ML made it possible for machines to learn from data and experience. By training algorithms on large data sets, machines were able to recognize patterns, make predictions, and make decisions without explicit programming. This led to a revolution in various sectors, including finance, healthcare, and e-commerce.

Deep Learning (around 2010)

‍Deep Learning, a subset of machine learning, came into the limelight around 2010. It uses artificial neural networks - inspired by the structure of the human brain - to recognize complex patterns in large amounts of data. With the increase in available data and advances in computing power, Deep Learning began to achieve impressive results in tasks such as image and speech recognition. This allowed machines to mimic human abilities at a level that was previously unthinkable.

Generative AI (Around 2020)

‍In the 2020s, the rise of Generative AI was observed. This form of AI goes beyond recognition and analysis; it can generate new information that wasn't present in the original data. With techniques such as Generative Adversarial Networks (GANs), machines can now create realistic images, texts, and other content. This has led to innovations in fields such as art, music and design, but also raises questions about authenticity and ethics in a world where machines can produce content that is barely distinguishable from the real thing.

Narrow AI vs AGI

It's crucial to remember that when we talk about AI, there are usually two main categories: Narrow AI and Artificial General Intelligence (AGI).

Narrow AI involves systems that are designed and trained for a specific purpose. This includes a self-driving car that navigates from point A to B or a forecasting model that helps a baker determine how many croissants to bake on Saturday. These systems are impressive in their specialty, but their scope is limited.

On the other hand, we have AGI, a form of AI that has the capacity to perform any intellectual task that a human being can do. It is a system that can mimic and potentially surpass all human cognitive processes. Today, AGI is still the domain of the future and science fiction. Most AI systems, including advanced models such as ChatGPT, fall under Narrow AI. For example, while ChatGPT can have advanced conversations, it won't spontaneously start a debate on Twitter.

Nevertheless, prominent think tanks and organizations such as OpenAI and Google DeepMind believe that the arrival of AGI may become a reality in the future. This transition to AGI involves numerous ethical and social issues, such as the impact on the economy, employment and the wider society. It is therefore very important that there are strict regulations for companies that are at the forefront of this technological evolution. However, it is important to emphasize that the vast majority of current AI applications, around 99.9%, are absolutely safe and fall under the Narrow AI category.

How to use AI

As companies begin to explore the power of AI, they can do so from both a short and long-term perspective.

1. Short Term Adoption - Using Existing Tooling

‍In the short term, companies can benefit from the existing AI tools and software that are already available. The idea is to accelerate, support and automate current business processes. A striking analogy is that a graphic designer doesn't have to develop Photoshop himself to use it and improve his or her work. In the same way, companies can use existing AI tools to support almost any business process. With the rise of generative AI, these tools are expected to be able to take over up to 40% of tasks in various departments - from customer service to finance and HR.

2. Long Term Adoption - Transformational AI

‍On the other hand, we have the transformational approach to AI. Here, companies go beyond just using existing tools; they combine their own data sets with customized algorithms and models to create innovative products and services. This approach is aimed at future-proofing the company, creating distinctive value and gaining a competitive advantage.

The Three Planes of AI Adoption

‍When considering AI integration, it's essential for companies to think at three levels:

• Individual Level: How can AI help me in my specific role?
• Team level: How can AI make my team more efficient and productive?
• Organizational level: How can AI contribute to the company's overall growth and progress?

By considering these layers, companies can take a holistic and strategic approach to AI adoption, leading to optimal results and benefits.

Examples of AI in various sectors

There is an AI solution for every business challenge and because the technology can be applied in almost any role and process, the possibilities are limitless - the question is where to start to get the most return. To give a few examples:

Transport and Logistics:

• Short Term: Using existing route optimization software to reduce fuel costs and accelerate deliveries.
• Long Term: Implementation of self-driving trucks and advanced predictive maintenance systems for fleet management.

Construction and Infrastructure:

• Short Term: Use of AI-driven planning software to determine the optimal order of construction tasks.
• Long Term: Use of autonomous drones and robots for inspection and construction work, and AI-driven models for sustainable infrastructure development.

Care:

• Short Term: Implementation of AI diagnostic tools that scan medical images for abnormalities.
• Long Term: Development of customized treatment plans based on genetic information and predictive models for disease outbreaks.

HR sector and Recruitment:

• Short Term: Use of AI-driven tools for resume screening and first job interviews via chatbots.
• Long Term: Implementation of advanced AI systems that analyze and optimize the overall employee experience, from onboarding to retirement.

Education:

• Short Term: Use of AI-driven tools for personalized learning paths and automatic assignment assessment.
• Long Term: Development of fully interactive, AI-based virtual classrooms and learning environments that adapt to each student's individual learning style.

Sales and Marketing:

• Short Term: Implementation of AI chatbots for customer service and use of predictive analytics for sales forecasting.
• Long Term: Using advanced AI to create fully personalized marketing campaigns and real-time customer behavior analytics for dynamic pricing and product development.

Concerns about AI

Artificial Intelligence, or AI, is often seen as a magic box, but in reality, it's based on math and statistics. An essential part of AI is the incredible computing power it requires. This computing power comes from data centers that consume a lot of electricity, raising questions about the sustainability of AI.

Another important aspect of AI is the required data. Data reflects reality and often includes the biases and biases we have as a society, known as “bias.” This bias can lead to adverse outcomes when integrated into AI systems, and is a source of major concern.

In addition to sustainability and bias, there are also fears about a future where AI will become smarter than humans, a concept known as superintelligence. Major companies in the AI sector, such as OpenAI, Anthropic, and Google DeepMind, believe that such systems could become a reality within the next five to ten years.

Another worrying development is the potential for autonomous weapons. The possibility of AI-driven weapon systems can lead to “Black Mirror” -like scenarios that raise ethical and safety issues.

The rise of AI also raises questions about social inequality. While some people will be able to adapt more easily to the AI-driven world and thereby become more attractive in the labor market, others may be left behind. What is the role of jobs in a world where automation is increasingly present?

It is crucial that we reflect on these issues. The European Union has already started introducing the “AI Act”, and the White House is looking at stricter regulations for AI. However, AI policy is still evolving. The future of AI has not yet been determined, and it is up to us to determine what role it will play in our society.

Conclusion

The impact of AI on our society is enormous and it will only grow in the coming years. As we continue to learn and evolve, we need to be aware of the ethical and moral challenges that AI poses. But one thing is certain: AI has the limitless potential to change the world!