What is artificial intelligence? The AI, simply put, is the attempt to transfer human learning and thinking to the computer, and thus give it brightness. Instead of being programmed for every purpose, an AI can find answers and solve problems independently.
The aim of AI research has always been to understand the function of our brain and mind on the one hand and to be able to recreate it on the other artificially.
Artificial Intelligence In Science Fiction And Reality
The dream of artificial intelligence is older than the computer itself – we know it from books and films, be it “Frankenstein’s monster” or artificially created people like the homunculus.
The term “artificial intelligence” has mainly come up in science fiction and usually means robots or computers that can think and act independently. Whether for good, like the android “Data” from “Star Trek” or for foul, like the computer HAL from the film “2001: A Space Odyssey”. In art, they are a means of asking questions about ourselves: what makes a human? What is intelligence?
However, when we talk about AI in today’s world, it has little to do with what we know from films and books. We only encounter AIs in a hidden way in real life – when new products are recommended to us on Amazon, when people are automatically recognized in photos, or when we chat with “Alexa” or “Siri” on our mobile phone.
Define The AI Term
So what is an AI? It’s difficult to say with certainty because there is no universal definition of artificial intelligence – because the term intelligence itself has not yet been clearly defined.
We’re trying to approach the term differently: In German, a distinction is often made between strong AI and weak AI when defining AI. Explained: Strong AI means what we know from science fiction. A machine that can solve problems of a general nature – any question you ask it. It is still pure fantasy and will remain so for decades or centuries.
On the other hand, we deal with weak AI in everyday life: These are algorithms – and that’s what an AI is, a very complex algorithm – that can answer specific questions that they have previously learned how to solve. An AI has no consciousness of its own and shows no understanding.
What Makes An AI?
In the following, we will only talk about weak AI since it is ultimately the only commercially relevant form today – we find weak AI in everyday life in our cell phones and computers.
What distinguishes an AI from a simple program? Typically, a programmer writes code in a language of her choice, consisting of a set of arbitrarily complex statements:
- If this, then that
- For example: when the user presses “Send, “send the email to server X
Such a system is also called rule-based. The programmer does not specify each step with artificial intelligence but writes an algorithm that can create these steps independently. As a rule, an AI does not report its code (even if there are already initial approaches here) but changes specific parameters within its code to find general patterns in data.
Why is that important? Because specific problems are so complex that it’s impossible to write code for them by hand. An example of this is the image recognition used in social media like Facebook: no programmer in the world can write a set of instructions that permanently recognizes what I look like, whether the photo was taken at night, at the beach, or on the car – In a rule-based system this would be utterly impossible because the programmer would have to know and be able to describe all possible images in advance.
A programmer is now teaching an AI how to recognize people but not how to remember me. The AI doesn’t know every picture of me either, but it can learn what I look like from several existing photographs and then apply this rule to new concepts and recognize me.
And not just with me, but with billions of faces in fractions of a second. THEREFORE, an AI can deal with previously unknown data, find patterns, or derive actions from them. It learns independently from the data available to it – but what it knows is determined in advance by the humans who design the AI. Humans program the AI, but it independently learns how to carry out its programmed task. AI is, therefore, far more powerful than rule-based systems, as they can – to a certain extent – react to previously unknown situations and learn from experience.
What Can An AI Do?
The possible uses of such AI systems are gigantic and not yet clear to most people. It will revolutionize our economy. The Federal Network Agency, for example, assumes that AI alone will create a value of EUR 430 billion by 2030, while the market study by Allied Market Research considers a global market size for AI technologies of EUR 1.5 trillion in 2030.
AI can extract information from data that a human could never grasp, for example, because it is too numerous or the underlying patterns are too complex.
Imagine if Youtube employees had to manually view every uploaded video and check whether it contained banned or stolen content. Five hundred hours of material is uploaded to the platform every minute. The group would need 90,000 employees alone who watch videos 8 hours a day without a break to keep up with the viewing! An AI manages this during the upload process, almost in real-time.
Artificial bits of intelligence like this is very good at capturing so-called unstructured data. These are, for example, images, videos, or sound recordings – data that cannot be easily searched because they do not have a consistent form in contrast to, for example, a table generated from sensor measurement data. While a traditional search algorithm (such as when you type CTRL+F on this website) can find an image’s title (a structured date), it cannot see whether Susie Mustermann is in the picture – that information isn’t written anywhere. It’s part of the image content. An AI can do that.
Of course, AI is also used to sort structured data and search for patterns. The current boom around AI takes advantage of unstructured data being much more common: it makes up about 80 percent of all data. With the rise of the Internet, Industry 4.0, and the massive availability of (cloud) storage, it has only been available in large quantities for a few years. Many companies do not even know what treasures of data they have and their value creation potential. With it, machine data, audio recordings of customer calls, or recordings of transport routes. You can read a few examples of this later. The massive availability of data in connection with the enormous advance in computing speed has led to AI being usable on a large scale in recent years.
The Google AI “AlphaGo” became famous, for example, which defeated the world’s best player in the board game Go in 2016 with game strategies that were previously unknown and have since changed the way people approach the game. A new version, MuZero, can even learn and optimize the rules of a play by itself. This ability to learn makes AI potentially applicable in many areas – and much more potent than its predecessors, which previously had to be reprogrammed for each purpose.
What Can An AI Not Do?
AI is not a general problem solver – not yet. She can process data exceptionally well and recognize patterns, but she cannot understand them. Artificial intelligence has no “common sense” – no common sense. It does not remember this if it comes to wrong conclusions due to insufficient data or poor programming (see section “Artificial Intelligence and Humans”). It can only provide answers to the specific questions it was programmed to answer.
Examples Of AI Projects
AI has long since found its way into our everyday lives. The example of facial recognition on social networks is one among many. Another is language assistants on our cell phones – we use Siri, Alexa, and Co. quite naturally in our everyday life. Translators like Deepl can translate our words almost perfectly into other languages in seconds.
When we surf the Internet daily, the advertisements we see are selected by artificial intelligence, which tries to display the most attractive product for us based on our interests and activities. We encounter these so-called “recommendation systems” online: Amazon, Google, Netflix, and Facebook. They are an effective system. As more and more media are vying for our attention, there is more to discover online than we can ever perceive in life. Computers, therefore, have to pre-select for us – and AIs learn over time to better understand us and play our preferences (against us).
But AIs are also finding their way into our everyday lives away from the online world. Robot hoovers clean our floors and use algorithms to recognize their surroundings. Navigation systems find the optimal route. Autonomous vehicles are making the most outstanding progress, accumulating millions of test kilometers on roads – even if they are still years away from widespread use. After all, Mercedes received model approval for autonomous driving on the motorway up to 60 km/hin 2021, making it the first manufacturer to achieve level 3 of 5 on the independent driving scale.
A few more concrete examples :
- The Bremen start-up ADD AI is working with the Werder Bremen football club to analyze reports from talent scouts using AI to find new football stars.
- Google ( Waymo ) is already testing the use of autonomous vehicles in practice – albeit currently with a driver as the last resort.
- Paypal uses AI to detect fraud attempts in the payment system.
- The Telekom AI “Tinka” processes 120,000 chat requests per month, it can solve 80 percent of all customer requests, and it refers to one-fifth of human employees.
Even if large corporations have mainly used AI, medium-sized companies can also benefit from it. One example is wind power: The icing of wind turbines is to be predicted in the PiB research project. The medium-sized Bremen wind farm operator, wpd wind manager, works here.
Different Types Of AIs
Many very different technologies researched over the past 70 years come under AI’s umbrella. The examples and methods described so far relate to a particular area of AI research, machine learning (ML). It stands for learning from experience. We have limited ourselves to this area so far since ML is the most relevant form of AI for companies in commercial use today, and a lot of the latest research comes from this, whether it is speech recognition (natural language processing using so-called transformers ) or image processing ( using Deep Neural Networks). More on this in our article on neural networks.
But there are also entirely different approaches. This includes so-called expert systems, which draw on a knowledge base compiled by experts to conclude using specific rules – they are more or less the opposite of “learning from experience.” The most famous example of an expert system is the chess computer “Deep Blue,” which defeated world chess champion Gary Kasparov, in 1997.
Both approaches are often classified into different categories – symbolic and sub-symbolic AI. A symbolic AI comes to results understandably by combining symbols (i.e., words, letters, numbers, etc.) according to pre-programmed rules to conclude. An example of this would be classical logic:
Symbol 1: “All men are mortal.”
Symbol 2: “Socrates is a man.”
Conclusion: “Socrates is mortal.”
On the other hand, a sub-symbolic AI does not arrive at a result by combining symbols and rules. On the other hand, it breaks down information into mathematical formulas and tweaks them until they produce the desired result. It is not possible to follow the development path directly from the procedure. This is experiential learning – machine learning.
Both AI approaches are not mutually exclusive – there are efforts to combine them or use elements of one in the other. More about this in: Understanding the difference between Symbolic AI & Non-Symbolic AI.
ALSO READ: Storage Solutions For Use With AI And ML
Use Artificial Intelligence In The Company
For companies, the use of AI for their processes can be attractive these days. For specific problems, they promise significant gains in inefficiency. Therefore, companies should ask themselves: What can I achieve with an AI? The first thing to look at is your data – what data already exists in the company, what could still be recorded? An AI can conclude them that were previously not possible – for example, because the analysis effort would be too high for humans or there was no way to get the correct answers – or nobody has even thought of generating data from them, for example, from an “old” machine that is not yet connected to a computer.
What companies can expect: once a job is found for AI, it will do it better than any human. Because it is not only faster, the error rate continues to decrease due to the constantly growing wealth of experience. According to the company, the Google AI “Lyna” (Lymph Node Assistant) can detect breast cancer in images with a probability of 99 percent, a value that doctors dream of.
It is essential to find a concrete use case because AIs are not (yet) available problem-solving machines. For example, one claim would be: “We want to check the quality of work parts from the assembly line in real-time using camera analysis, without resorting to manual random samples.”
Like all far-reaching innovations, the successful implementation of AI in a company takes time. The return on investment for a project is between 12 and 18 months, estimates Roland Becker, Managing Director of the Bremen-based AI expert ADD AI. Appropriate knowledge is also necessary for a project to be successful, and the excellent quality of the available data. In addition to hiring your experts, working with cooperation partners in research projects ( such as with the Bremen BIBA ) is particularly useful for medium-sized companies. They gently introduce the topic and enable you to get to know the new technology with relatively little use of resources.
Because the training of intelligent networks requires a high level of computing power, which can be achieved either by investing in or renting cloud capacities – a partner who already has the ability for this makes it much easier and cheaper.
So – Do Medium-Sized Companies Now Absolutely Have To Rely On AI To Survive?
Medium-sized companies usually find it challenging to adapt to new technologies quickly. There is a lack of the resources of large corporations for experiments and the agility of start-ups without running costs. It’s the same with AI: According to a survey by the Bitkom industry association, most companies with fewer than 500 employees have refrained from investing in AI. The reasons are a lack of human resources and time and often competitive pressure. Many companies are still waiting.
So is it better to wait? The answer is clear: yes. AI technology is still young, although it has been researched since the 1950s. The computing capacity has only been sufficient to operate AIs commercially for a few years. It is new territory, and a successful medium-sized company that runs without it today will still run without it tomorrow.
For small businesses, investing in artificial intelligence is a gamble. Therefore, the first question should be: How could an AI increase my sales? How could an AI reduce my costs and improve services? How can my customers benefit? It helps to deal with the technology to get an overview of the possibilities. Free information offers, such as those of the Mittelstand 4.0 centers in Germany, help accumulate knowledge. If a use case is found, an idea for use, local partners, and funding help implement it.
Although large cloud companies such as IBM, Google, or Amazon also offer AI solutions, these can quickly be oversized, especially since it still takes experts to implement them successfully. And skilled workers are rare, especially in the field of AI. Anyone who currently sees no purpose for an AI should stay on the ball: One day, it will come to the point that competitors will bet on it, and it will be too late by then. And at the rate at which AI is currently evolving, that time will come sooner rather than later.
At the same time, the costs and resources required for using AIs are falling rapidly. So-called frameworks have been around for several years and provide the essential tools to set up your own AI networks quickly – TensorFlow and PyTorch are the most common. This means that even small companies can set up AIs – the Bremen 5-man company INnUP is a perfect example of this. At the same time, work is being done on systems that also enable laypeople with no experience in programming to use AIs.
Artificial Intelligence And Humans
Like many new technologies, AI is fueling fears. A famous study by the University of Oxford analyzed in 2013 that 47 percent of all US jobs are at risk from automation, a significant proportion of them from AI. Such numbers stir up fears that lead to tangible actions: Waymo, the Google subsidiary for automated driving, reports that its test vehicles were repeatedly attacked with knives and stones. So is AI a danger to humans? A bottom survey paints a mixed picture: 62 percent of Germans see AI primarily as an opportunity, 35 percent as a danger. Also, a survey among managers showed that 42 percent of them observed reservations in the workforce.
The truth lies somewhere in the middle. AI will no doubt take over human labor, and when it does, it will be in its entirety—that is, a human will no longer be needed for this one specific task. These are primarily tasks with minor fun factors and are monotonous and repetitive: watching surveillance videos, answering formal inquiries on the phone, and searching through documents.
However, in many areas, AI will continue to play an assistant role for the foreseeable future. They will help doctors come to the correct conclusions or be part of a work chain – for example, in the field, where agricultural robots can take over partial tasks autonomously or semi-autonomously. At the same time, humans still carry out other studies.
At the same time, however, new jobs will be created, supported by the innovative AI business models. People then have more time to use their labor for new tasks because they work with AI. In this way, lawyers could spend more time with clients instead of spending hours trawling through files. It is also clear that more education is needed to prepare people for their new tasks and give them the skills to work with AI systems.
A study by theWorld Economic Forum of 15 industries in 26 economies assumes that 97 million new jobs offset 85 million jobs displaced by computers.
And, to be honest, we don’t have a choice. Because AI has long since found its way into everyday life, almost everyone already uses it, albeit unconsciously – whether on cell phones, bank transfers, or navigation. It will be some time before we see AI everywhere. Still, that time will come sooner rather than later because once an area benefits from AI, it will have massive advantages over its human counterparts, driving them out of the market.
Nonetheless, it’s essential to talk about it and ask yourself where the ethics are in the machine. This is not just about responsibility (“Who is to blame if the machine causes an accident?”), but also how we want to organize work in the future.
The Natural Stupidity In Artificial Intelligence
AIs are manufactured and subject to a natural problem: an intelligence that imitates humans are also subject to their mental limitations. One of them is bias.
An Example: In 2014, the AI experts at Amazon developed an AI that automatically evaluated and sorted application documents. They trained the neural network with applications from the past ten years to do this. When the AI was introduced, they found that the algorithm only selected those from men among new applications. Reason: Among those hired earlier, there was an above-average number of men, standard in the tech industry. From this, the AI created the rule: Hire only men. (Source) The error lay in the selection and preparation of the data. Amazon ultimately rejected the experiment and continued to search through applications manually.
The example shows that when designing artificial intelligence, people must attach great importance to selecting representative data – and are aware that they may already be biased when choosing and processing the data. This dilemma is difficult to solve and must be considered in any AI design. This is one of the reasons why it is worth taking a look from the outside, working with a partner and an expert in the field of AI.
After all, every AI is programmed by a human – and we know where our intelligence begins and ends.
Finally, here is a summary of what artificial intelligence is:
- AI is the attempt to transfer human learning and thinking to the computer.
- Strong AI, i.e., general problem-solving machines, belongs in the realm of science fiction. Weak AI is being used more and more today, whether in mobile phones, websites, social media, or self-driving cars.
- AIs are valuable wherever there is a lot of data to analyze and search for patterns.
- Machine learning is currently the commercially most crucial sub-area of AI.
- AIs need data as a basis, which can be images, videos, or sounds in addition to numbers.
- AIs can process data better, more accurately, and faster than humans, but they cannot understand it.