Introduction
Since computers were first invented, IT technology, especially software and artificial intelligence technology, has been developing at a tremendous pace. Now, in fields such as Go and chess, computer software has advanced so much that it cannot be compared with humans anymore. In 2015, an AI called AlphaGo appeared and shocked many people when it overwhelmingly defeated a professional 9-dan Go player with a score of 5-0 without any handicap for the first time. Since then, AlphaGo has won all 74 matches against professional players except for one against Lee Sedol. Since the advent of AlphaGo, people’s interest in AI has increased, and lots of people are worried about the development of AI. Still, people believe that AI cannot catch up with humans in fields that require creativity, such as art and music. However, recently, it has been controversial again after the emergence of AI, which draws high-quality pictures just like artists..
Controversies Over New Technology
Recently, one artist submitted a picture at the Colorado State Art Competition, and this picture won first place and the prize. However, after a while, it was revealed that his painting was drawn by AI (Kuta, 2022).
[Jason Allen (2022) Théâtre D’opéra Spatial. [Online Image] Available from: https://www.westword.com/arts/artificial-intelligence-art-colorado-state-fair-controversy-14963306. [Nov 22, 2022]
(Picture drawn by AI, Midjourney)
An artist raised a problem about using AI to participate in a competition, which sparked controversy and multiple opinions.
AI began to threaten even the realm of creativity, which was believed to be impossible to do with technologies. These incidents raise questions about whether AI can really replace humans.
Can AI replace humans in the field of art?
People are debating these questions a lot nowadays. Can AI replace human artists? Furthermore, can robots replace humans in every area.?
First of all, even if the drawings drawn by current AI outperform those drawn by human artists, AI drawings are just pictures imitated based on the drawings of many human artists. In other words, unlike humans, AI did not create something from nothing but rather analyzed and imitated existing data. This is the limitation of current AIs.
However, if AI becomes highly developed and it is possible to implement a human brain structure in a computer that is very similar to a human brain, a robot that thinks the same as a human might be created. If so, in the future, AI might replace humans even in fields that we believe are unique to humans.
What is Big Data?
“Big data” refers to numerous data sets that go beyond the limits of existing software for data collection, management, and processing. Big data, one of the ten emerging technologies, has a variety of characteristics, one of which is that its size is constantly changing in real-time (Breur, 2016).
The common characteristics of big data can be defined as “3V,” which is volume, velocity, and variety. Volume denotes the amount of data, velocity denotes the rate at which data is created and processed, and variety denotes the variety of data types. Recently, veracity and variability have been added since the reliability may be lowered as there is a lot of data, as well as for the utilization of the data (Anil Jain, 2016).
There are numerous examples of big data usage, and the COVID-19 situation is one of them. Big data was utilized to minimize the spread and impact of the disease. Governments in South Korea, Taiwan, and Israel adopted it early to prevent and slow down the virus pandemic.
However, big data does not only have advantages. There are cases where these data have been used without permission in research, and they might infringe on individual privacy. Therefore, a cautious attitude is required when using big data.
What is deep learning?
Deep learning is one of the machine learning methods based on artificial neural networks. Through deep learning, the computer learns and imitates the way humans understand certain types of knowledge. Artificial neural networks attempt to imitate the way the human brain thinks through a combination of numerous data inputs. By analyzing countless pieces of data, the computer can recognize, classify, and describe objects accurately based on its experience (IBM Cloud Education, 2020).
In fact, although there was a deep learning concept in the 1980s, it did not shine due to many hardware and software limitations. In 1989, one team introduced a deep neural network to recognize zip codes in the mail (LeCun, 2015). However, the time it took to train the neural network was so long that it was considered unrealistic.
Deep learning was revived in the 2000s as software weaknesses were overcome and graphics processing units with powerful performance were released. For deep learning, the GPU is used since it is exclusively designed for parallel computation. Multiple expensive graphics cards with high VRAM capacity and performance, such as QUADRO, RTX TITAN, and RTX 3090, are used.
Using artificial intelligence
With the development of software and hardware, a lot of tools using AI have recently appeared.
- NVIDIA Canvas
In 2021, Nvidia launched Nvidia Canvas, a tool that uses AI to convert simple user drawing into realistic images instantly.
[Nvidia (2021) Nvidia Canvas demonstration. [Online image] Available from: https://www.nvidia.com/en-us/studio/canvas/. [Nov 23, 2022]
NVIDIA Canvas AI Tool Demonstration
This tool saves a lot of time since it uses AI through the tensor core of the graphics card to allow the user to customize the picture. This image, drawn by AI, is of such high quality that it is believed that it is a real photograph.
- Midjourney
Midjourney is an artificial intelligence program that generates paintings based on user-written text or description, which was announced in 2022.
[MantaRay (2022) Midjourney art example: Asian Warrior Chief [Online image] Available from: https://tokenizedhq.com/midjourney-examples/. [Nov 23, 2022]
Midjourney AI Prompt Demonstration
Anyone can use this program to create art illustrations for free. The users just need to input a few words, and the AI draws the appropriate pictures for them.
The Future of the Art
For the time being, AI cannot completely replace humans due to its limitations. For example, AIs are struggling with rendering human hands. Also, the paintings drawn by the AI are not perfect and have defects. Therefore, AIs are likely to be used as a supporting tool for a while. However, if these deficiencies are resolved through sufficient learning from more data and software improvements, there is a possibility that AI will completely replace humans even in artistic fields in the future.
References
- Anil Jain (2016). The 5 V’s of big data – Watson Health Perspectives. [online] Watson Health Perspectives. Available at: https://www.ibm.com/blogs/watson-health/the-5-vs-of-big-data/ [Accessed 23 November. 2022].
- Breur, T. (2016). Statistical Power Analysis and the contemporary ‘crisis’ in social sciences. Journal of Marketing Analytics, [online] 4(2-3), pp.61–65. doi:10.1057/s41270-016-0001-3.
- Burns, E. (2021). What is deep learning and how does it work? [online] SearchEnterpriseAI. Available at: https://www.techtarget.com/searchenterpriseai/definition/deep-learning-deep-neural-network [Accessed 23 November. 2022].
- IBM Cloud Education (2020). What is Deep Learning? [online] www.ibm.com. Available at: https://www.ibm.com/cloud/learn/deep-learning [Accessed 23 November. 2022].
- Kuta, S. (2022). Art Made With Artificial Intelligence Wins at State Fair. [online] Smithsonian Magazine. Available at: https://www.smithsonianmag.com/smart-news/artificial-intelligence-art-wins-colorado-state-fair-180980703/ [Accessed 23 November. 2022].
- LeCun, Y. (2015). Backpropagation Applied to Handwritten Zip Code Recognition. [online] web.archive.org. Available at: https://web.archive.org/web/20150529000503/https://www.ics.uci.edu/~welling/teaching/273ASpring09/lecun-89e.pdf [Accessed 23 November. 2022].