AI Network
Published in

AI Network

You can study AI without coding

Interest in artificial intelligence is certainly hot these days. Although attention is peak, there are fewer people who have actually looked inside. Questions are bigger than conviction these days. As a result, extreme questions such as “In the future, artificial intelligence replaces human jobs…” and “Is the day when artificial intelligence dominates the world really coming as Elon Musk said?!” is common. Most of the thoughts of non-publicists and non-developers are probably staying here, even though they see the content recommended by artificial intelligence, encounter artificial intelligence-based advertisements, and live in platforms with artificial intelligence here and there.

To understand Artificial intelligence, you have to learn and study.

When the old car first came out, there were no roofs and windows. In winter, if there were no thick winter clothes, and if there were no umbrellas or raincoats on rainy days, people had to see failure in the car. There was no way for cars to run. That’s why people had to carry shovels and saws when they were running on a rough road. It is only after Gasoline cars came that people began to study inconvenience in cars. It’s in a similar vein. In order to turn the question of artificial intelligence into conviction, it is necessary to ‘populate’ artificial intelligence. It is necessary to experience “Ah, this is what artificial intelligence is like” by getting hit by the snow, getting rained on, and sometimes making a path that doesn’t exist. In order to do so, everyone must learn and study artificial intelligence. There is no area where the cliché of ‘To see is to believe.’ fits this well.

DILIVE: You can study AI without coding?

These days, even if they are not developers, they often study coding, but the reason why ‘artificial intelligence’ is still a distant story for the majority of the general public is ‘coding’. We can’t understand artificial intelligence in the frame put on by someone, blocked by barriers to what we believe is accessible in the developer’s domain. In that sense, ‘Nocode’ that does not require coding is drawing attention these days. More people can become manufacturers and participate in the market if it becomes easier to make something on the Internet, just as ‘one-person media’ now makes broadcasts that used to be made by only a few stations. The world has come where ordinary people, not 1% of engineers, can implement ideas on computers. Just as anyone can simply edit through programs already created when dragging animation functions on PowerPoint, solving complex formulas on Excel with a few buttons, and editing videos, a new wave of ‘Nocode’ is flowing into artificial intelligence.

The third topic of Data Intelligence Live at the clubhouse in South Korea, ‘Can you study AI without coding?’ was a forum for checking and informing this trend.

If you can study artificial intelligence without coding, it will open a horizon for many people to understand artificial intelligence. Kim Dae-sung, PM of Common computer, who are creating a Nocode community with this meaning, Kim Byung-hak, an artificial intelligence developer and data intelligence podcast operator Kim Jin-young conducted the discussion. It was joined by about 100 participants.

Kim Jin-young is a data scientist. He is currently hosting a data intelligence podcast and is the author of “Hello Data Science”. Kim Byung-hak is currently leading research and development of machine learning technology at Alpha Health, an AI startup in Silicon Valley. Kim Dae-sung,PM of Common computer worked for Naver Labs’ UX designer and Samsung Electronics’ wireless UX designer.

Should we learn AI?

The first area to be discussed was the area of ‘education’. Artificial intelligence is still the domain of scholars and developers who study AI. Should we learn AI? If you have to learn, what is the purpose and what you should think first of all about the possible means and direction of education.

The most intuitive reason to know artificial intelligence is to recognize trends and possibilities of technology development. Even if it does not go into deep algorithms, it is necessary to have enough skills to learn and use necessary technologies on time.

Kim Jin-young pointed out that in foreign countries, Google’s trend search volume shows a lot of things about machine learning and data science other than deep learning. It means that interest is evenly distributed. In the case of Korea, it is mainly focused on deep learning-oriented communities and education markets. Machine learning is to analyze data using algorithms, learn from analysis, and make judgments or predictions based on what we have learned. The goal is to learn how to perform tasks by learning the computer itself through a large amount of data and algorithms rather than directly coding the software for decision making. Deep learning is an advanced form of artificial intelligence in artificial neural networks. Data is learned by utilizing information input and output layers similar to neurons in the brain. Although deep learning is showing good performance in many areas, it was noted that data analysis, which scrutinizes the results of traditional machine learning or learning, and algorithms, is still an important topic.

The case of education presented by Kim Byung-hak was also impressive. MIT Media Lab, which has led artificial intelligence research since the 1960s, has created a curriculum through various activities so that elementary and middle school students can regard artificial intelligence as one of the opinions. In addition to the latest research trends related to artificial intelligence, it contains project-based activities and software tools for children to learn artificial intelligence through collaboration. It will also provide channels that can be accessed directly with other AI education resources such as Google Experiments. MIT emphasizes the mixing of plugged and unplugged learning approaches. In other words, children are educated about what AI is like and what the social impact is without using a computer. Google also has Teachable Machine, which allows users to experience artificial intelligence without coding. If education using these tools is expanded to the base of society, it will greatly help popularize and publicize artificial intelligence.

Kim Dae-sung expressed the need for a change in perception of machines.Machines are no longer just providing the correct answer or set function. Results are based on statistical probabilities. If the paradigm in the past was Human Computer Interaction, it is now changing with the interaction of human and artificial AI Interaction. The development of computers and the Internet has affected not only engineers but also the entire industry. Artificial intelligence may be the same. This is why everyone who is possible should learn artificial intelligence. Some said it would be necessary to accept the results of AI as a ‘opinion.’

What data do you need to use artificial intelligence?

The problem lies in what kind of artificial intelligence is used in practice. A task that always inevitably follows artificial intelligence is ‘data’. Big Data requires massive amounts of intangible and intangible resources to manage them. To be used in practice, artificial intelligence based on ‘easy-to-acquire data’ and ‘applicable data’ can be the answer. Examples include MS Azure ML Studio using MS Azure, an open cloud platform that can build, deploy, and manage applications in Microsoft data centers around the world. Microsoft provides years of machine learning research as a simple and easy-to-use single cloud service, making data accessible to first-time users at low cost and no additional software or hardware purchase required. Another advantage is the development environment where experiments can be conducted using various algorithms, and the trained model can be distributed as an API and easily utilized.

Why did you start the Nocode AI methodology?

Kim Dae-sung’s personal experience was impressive. It will also be something that many non-publicists sympathize with. Everyone talks about artificial intelligence, and our work is definitely an area that is directly and indirectly connected to artificial intelligence, and sometimes we feel like something is left out. In particular, Common Computer is a startup that creates infrastructure and platforms for AI development and services, so it requires understanding and knowledge of the deep part. Learning AI as a non-publicist is not easy (but when you think about the Internet, it’s not impossible either). Some of you may remember a website called Wix. Wix, which does not create web pages by coding them one by one, but can easily create a homepage without development knowledge, has also achieved great growth through the untact wind.) It was also this point that they realized that they needed an AI learning methodology that anyone could learn from a limited market targeting developers.

What is a representative of the Nocode AI solution?

Nocode is an area that is already actively applied in the field. One of the most representative is the Nocode platform called Unqork in the U.S., which has already attracted investment from Goldman Sachs, Alphabet and Blackrock and has been recognized for its $2 billion corporate value. You can create apps or websites in weeks or days.

In terms of data, the Corona API or company’s bankruptcy prediction system, created using Azure ML studio and Parabola.io, and other web app platforms such as determining coronic app using deep learning, video interview simulation app, and others are actively used.

It is also curious about the actual application cases that are being carried out closer. Under the leadership of Daesung, more than 100 people from Ewha Womans University, Common Computer, and the community are studying hard to make actual application cases. Kim Dae-sung is in charge of a product called Ainize.ai, which allows users to conveniently use AI as an API, without the need to develop new AI, and when he introduced it to the study, more than half of the participants were able to implement AI services. An OCR-based service called Vday, which is actually about to be commercialized, is led by a non-developer designer.

What’s the possibility of a Nocode?

In fact, everything starts with ‘desperation’. Economical thinking, which uses limited resources efficiently in busy daily life, is a technology of survival. A place where ‘low cost and high efficiency’ does not work well is the world of dev. Development is not going as fast as we thought. New technical problems are constantly popping up. Modification and repair are also essential. In many cases, it is difficult to rewrite the code that can only be understood by developers. Some say that it is difficult to find developers. No code can solve these problems quickly and cheaply. Some suggested that there is a big need in terms of being able to sketch and start with no code rather than putting developers in from the beginning. No code will be first applicable to industries that lack developers and experts. The problem is that there are not many cases that have reached the production level yet. Even if you don’t code, you should study at least. Continuous investment and education are inevitable in this regard. It is encouraging to hear that the recruitment phrase for Nocode developers has also appeared recently.

How can we overcome the limitations?

For South Korean who have never been in the states, one of the hardest things about graduate school was that all the subjects were written, speaken in English. It is difficult to study difficult international law, but all kinds of difficult terms were written in English, so it took twice as long to study. In the same manner, it would certainly be helpful if we could learn artificial intelligence in a language that we can understand. But it is accompanied by limitations. Reading a book translated into Korean can, of course, help you understand it right now, but you should eventually read the English application again because the world in which it is actually used is an English-based world. While no code may be sufficient for patterns already used, new innovations come from code and mathematics.

Nevertheless, the bright future of Nocode comes from the power of ‘convergence.’ In a social atmosphere divided into developers, non-developers, liberal arts and natural sciences, the self-mixed words ‘Sorry for study humanites, liberal art’ are heard here and there, and. The phrase ‘developer is the best’ is also common. The world is not divided into liberal arts and natural sciences. The world has not developed only by the power of science, and the world does not work only by the power of philosophy or money. Just as artificial intelligence requires ethics, science requires philosophy and vice versa.

Kim Min-hyun, CEO of Common Computer, explained the ‘developer’s thought process’ in the middle. Developmental thinking that needs to be seen separately as if it is building Lego is also very important for technology companies. As he said, the blind spot lies in the risk that one completed picture may not appear altogether. In the end, it is necessary to have a humanistic perspective that is familiar with writing long but well-organized writings on blank paper. I think the real power that no code can bring lies in ‘convergence.’

Do you have any plans regarding the Nocode AI community or activities in the future?

This is where the Nocode AI community and activities are more expected. Like the Teachable Machine made by Google, the company plans to make efforts to create practical cases of Nocode products and services that can be easily used by the general public. At the Common computer company level, artificial intelligence experts and domain experts are working together to easily utilize the artificial intelligence platform through cooperation with large companies(One of the Korean big-tech).The meeting between data scientists, real artificial intelligence developers, and non-deveploper was both exotic and meaningful. I hope that more people will be able to be together through future broadcasts.

This article summarizes the contents of the #DILIVE clubhouse broadcast co-hosted by PM Kim Dae-sung of Common Computer. Please tell me the source when you share it. Please click the link below to come to the Nocode community!

If you want to know more about us,

--

--

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store
AI Network

AI Network

AI Network official account. Please contact me here. info@ainetwork.ai