Artificial Intelligence (AI) is currently known for its use in super-intelligent humanoid robots. Usually, artificial intelligence is implemented behind the scenes as an algorithm that processes large amounts of data to perform a range of mundane tasks more efficiently than a human.
Although most of us do not drive a self-driving car (yet) or use a humanoid robot, our daily lives are increasingly influenced by AI systems that recognize sound or images or analyze our online behavior to protect us from credit card fraud or show us appropriate advertisements on the web.
It is therefore no wonder that the boom in this type of artificial intelligence is being driven by internet giants such as Google, Microsoft, Amazon, Facebook, Baidu and others at the forefront and that there is currently a resurgence of startups in this industry:
It is also no surprise that AI and automation are found high up in Gartner’s „2015 Hype Cycle for Emerging Technologies.“ As is the self-driving car at the top of the curve – the „Peak of Inflated Expectations.“
2015 Gartner Hype Cycle for Emerging Technologies.
Gartner classifies “Autonomous” as the sixth and final step of a company on its path to digitalization. Some emerging technologies are identified as particularly relevant: self-driving vehicles, bioacoustics, biochips, brain-computer interface, digital dexterity, human augmentation, machine learning, neurobusiness, people-literate technology, quantum computing, smart advisors, smart dust, smart robots, virtual personal assistants, virtual reality, volume displays and holographic displays.
As usual, what follows the hype is disillusionment. It is difficult to imagine that all of these emerging technologies from Gartner’s Hype Cycle will make it into the mainstream. However, according to Gartner, there is enough evidence for AI to become established in companies.
Why AI is taking off now:
To solve the problems that come with adopting AI, you first need massive amounts of computing power – something that has been dictated for the past few decades by Moore’s Law, which codifies the observation that if you shrink the size of transistors on an integrated circuit, you can fit more of them in a space. Although the number of transistors per chip is increasing thanks to smaller feature sizes (we’re now at 14nm, with 10nm and 7nm on the roadmap), metrics like clock speed and single-thread performance haven’t changed much since 2005 due to heat and power consumption.
Bild: ‚Efficiency and Parallelism, The Challenges of Future Computing‘ (Bill Dally/Nvidia)
In order to promote AI, other technologies are used that increase computing power. Computational operations can be parallelized in the cloud. By using any number of virtual computers, a calculation can be easily distributed and processed simultaneously by theoretically any number of computers.
Another development is the outsourcing of computationally intensive work to graphics chips (GPU). These can calculate neural networks particularly quickly and efficiently. Neural networks are currently the focus of AI research and are being heavily promoted by tech giants such as Google, Facebook and others.
Below is a comparison of a GPU calculation compared to a conventional CPU.
Aside from computing power, the other driving force behind the AI boom is the availability of cheap storage. The equivalent of Moore’s Law for storage is called Kryder’s Law. This law declared in 2005 that the areal density of hard drives would more than double every two years. This would then lead to similarly exponential increases in capacity and reductions in storage costs per GB.
AI in the company
According to a Deloitte report, the new benefits of technology have always eliminated some jobs and created others – and this trend will continue.
This is also the view of another report from Forrester in August 2015. They calculated that automation will lead to a loss of 9.1 million US jobs by 2025. This number is nowhere near the 69 million mark that many experts have predicted. However, automation should not be viewed as the „enemy“. Automation gives people time for more creative activities – a domain that machines cannot penetrate so quickly.
Currently, it seems that the fear of losing one’s job arises at the lower end of the income scale, as the YouGov study in the USA shows:
New start-up companies, for example, are contributing to certain jobs becoming rarer or even disappearing altogether in the future.
Startups like Narrative Science, whose advanced text generation platform „Quill“ can search through many databases to apply business rules to produce content that is clearly different from and better than content that a human could write. The startup Kensho is also swapping humans for machines: their new search engine for the financial industry, which analyzes and evaluates statistics and answers complex questions about finance in simple English, is more effective and faster than experts on the stock market can be.
AI is even needed in medicine. IBM’s MSK-trained Watson for Oncology system, for example, can search for treatment options from a large database for each patient whose information is stored in the system.
At the moment, these types of systems are more likely to empower people than replace them, as stated in Forrester’s report (The Future of Jobs, 2025: Working Side By Side With Robots). This opinion is supported by a Narrative Science study of 200 managers and CEOs on the topic of artificial intelligence and big data in the company. 80% of respondents believe that AI improves work performance and creates new jobs.
AI is classified by respondents as “technology that thinks and acts like humans” and is widely used in their companies through speech recognition and as a response system.
Quelle: Narrative Science
OUTLOOK
AI, primarily in the form of deep-learning algorithms running on powerful graphics processors, is already widely used in companies that are becoming increasingly data-driven and digitizing. This pays off for a company through the considerable price-performance ratio that AI brings and is characterized by new, highly developed algorithms that make work easier.
Opinions are divided as to how big an impact AI will have on the job market, but one thing is certain: AI will continue to evolve and play larger and more significant roles in our lives.
