In March 2023, Singapore’s Ministry of Manpower revealed its inaugural Shortage Occupation List, a condensed overview of in-demand jobs that local talent alone cannot fulfil. It comprises 27 occupations across industries. Nearly half of these roles are in technology, including within Data Science and Artificial Intelligence.
While AI has gained immense traction and is influencing various areas of our lives, it is no secret that building AI models and algorithms can be challenging and time- consuming. After all, it is a complex and technical process that requires specialised analytical knowledge and expertise.
Accessibility of AI
Imagine you are scrolling through your feed on Instagram or TikTok and suddenly stumble upon a life hack so simple yet useful, it makes you wonder how you have never known about it before. This is how the discovery of so called no-code AI tools makes many people feel.
No-code AI tools eliminate the need for specialized skills, allowing non-technical users to build and deploy AI models, without any requirement for coding or programming. Where ChatGPT would respond with “It is hard to predict…”, these tools transform user- specific data input into tailor-made AI models and predictions.
The idea of making AI accessible to everyone, regardless of their technical background, includes small and medium-sized businesses that may not have the resources to hire a dedicated team of data scientists or machine learning engineers. With user-friendly, drag-and-drop interfaces, no-code AI tools allow even those without technical skills to harness the power of AI predictions.
AI Education for the Public
AI’s relevance for a multitude of areas has led to its inclusion in a range of educational curriculums. Relevant courses can roughly be classified into two categories:
- Introducing the public to the fundamental concepts of AI, such as uses, risks, and limitations, however without enabling them to create actual models or predictions.
- Teaching technical skills in programming languages like Python or R for those ought to become specialists in the field, driving AI development for others.
No-code AI software bridges the gap between these two categories. A variety of such tools is available on the market, including Orange Data Mining (taught by AI Singapore), DataRobot (considered a favourite by many), and Obviously.AI (one of the newer players, with an emphasis on its intuitivity).
So, what can no-code AI tools do for you?
In a nutshell, AI in general can use existing data to classify or predict unknown cases. In the context of Singapore, for example, AI might analyse a table of historical HDB resale prices to determine how factors like square footage, construction year, and neighbourhood affect the final price, and then make future predictions based on that analysis.
With no-code AI tools automating the entire AI development process, from data preparation to model training and deployment, users simply need to upload data and let the tool work its magic. In an initial quality analysis, the tool may flag missing or inconsistent data entries for consideration, and then proceed to run tens of thousands of algorithms and parameters in the background, recommending the ones with the most realistic outcomes and predictions for deployment.
Simplicity applies to both input as well as output: A result may, for example, be a user- friendly, shareable online form to experiment with. In our HDB resale price scenario, users could input any random construction year, square footage, and neighbourhood of a flat, and the tool would generate a real-time AI-driven price estimate. Remember: We got all this by doing not much more than to upload an Excel table into the tool.
The Future of No-Code AI
As the technology continues to advance, we can expect to see more businesses adopting no-code AI tools to gain a competitive advantage and make better-informed business decisions based on data-driven insights. Tasks that once demanded significant time from skilled professionals can now be accomplished by a broader user base and in a fraction of the time.
Depending on the specific tool used, available functionalities may not only include classification, regression and time series forecasting problems, but even natural language processing, image recognition, hyperparameter tuning and more. Additional resources are available via the following links:
About the author: With a German background, Andreas Herpich relocated to his dream country Singapore in 2018. Apart from assisting clients in digital transformation, data analytics and process automation, he is an active advocate of empowering everyone to utilise the power of AI through no-code tools.