{"id":1606,"date":"2020-08-03T13:15:01","date_gmt":"2020-08-03T06:15:01","guid":{"rendered":"https:\/\/www.pacificprime.co.th\/blog\/?p=1606"},"modified":"2020-08-03T14:14:31","modified_gmt":"2020-08-03T07:14:31","slug":"ai-in-thailand-what-you-should-know","status":"publish","type":"post","link":"https:\/\/www.pacificprime.co.th\/blog\/ai-in-thailand-what-you-should-know\/","title":{"rendered":"Artificial Intelligence in Thailand: What you should know"},"content":{"rendered":"

Artificial intelligence or AI can often conjure up dystopian images of robots taking over the world. While the reality may be a far cry from our perceptions, organizations implementing AI solutions are still not able to take things lightly. Global advisory and research firm Gartner believes that\u00a0AI needs to be strategically leveraged<\/a> for effective deployment, so in this Pacific Prime Thailand<\/a> article, we\u2019ll tell you what this means for your organization in the context of an increasingly digitally-oriented Southeast Asian nation.<\/p>\n

\n

\"Corporate<\/a><\/p><\/div>\n\n

The centrality of big data<\/h2>\n

Coined in the mid-1950s by John McCarthy, AI is \u201cthe science and engineering of making intelligent machines\u201d. This means using computers to do things that require human intelligence.\u00a0In other words, it involves creating algorithms to classify, analyze, and draw predictions from data. There\u2019s also an element of acting on data, learning from data, and improving over time.<\/p>\n

While AI is a broader concept to make intelligent machines, machine learning (ML) is a subset of AI that enables machines to learn from data without being directly programmed. So, within these concepts, there is a centrality of data. Dr. Mark van Rijmenam, a globally recognized expert on big data and AI, says in an interview<\/a> that:<\/p>\n

\u201cThe datafication of your organizations opens a whole new line of possibilities, which you should approach from an out-of-the-box perspective. Only then will you be able to come up with real added value for your organization.\u201d<\/p><\/blockquote>\n

Implementing AI-based solutions<\/h2>\n

Gartner\u2019s 2020 CIO Agenda Survey<\/a> found that leading organizations plan to double the number of AI projects within the next year, with more than 40% aiming to deploy solutions by the end of this year. Organizations will struggle to scale AI pilots into enterprise production, limiting its ability to realize AI\u2019s full potential. The following gives five tips<\/a> to overcome the challenge:<\/p>\n

1. Refine AI model<\/h3>\n

In order to accelerate AI pilots into production, organizations will require specific infrastructure resources that will be able to evolve alongside the technology. Therefore, AI models must be refined periodically for high success rates. For example, these could involve standardizing data pipelines or integrating ML models with streaming data sources to deliver real-time predictions.<\/p>\n

2. Collaboration between business and IT teams<\/h3>\n

A challenge in leveraging AI techniques like ML or deep neural networks (DNN) in edge and internet of things (IoT) environments is the complex nature of data and analytics. It will be necessary to have collaboration between the business and IT teams, so that you can proactively plan and provide ready solutions when new business needs emerge.<\/p>\n

3. Simply ML techniques<\/h3>\n

The early AI adopters used ML solutions to deliver business value. This typically involved traditional statistical machine learning, but slowly progressed to use more advanced techniques with deep learning to grow the impact of AI. However, it is important to go beyond the AI hype and focus on the business problem at hand. Often, simple techniques are better.<\/p>\n

4. Utilize cloud service providers<\/h3>\n

AI deployment processes can be simplified through the strategic use of cloud technologies like containers and serverless computing, as this can allow ML models to serve as independent functions, reducing costs and overheads. A serverless programming model is especially appealing in public cloud environments due to its quick scalability.<\/p>\n

5. Embrace AI-augmented automation<\/h3>\n

As organizations have to manage increased data, there will be more ineffective problem prioritization – worsened if business and IT teams are not in sync. AI-augmented automation can help IT teams position themselves to have more effective partnerships with peripheral business units, resulting in higher IT productivity with greater agility and scalability.<\/p>\n

Where do we currently stand<\/h2>\n

Data-driven agendas are part of Thailand 4.0 <\/a>– an economic model that progresses from past developmental strategies that emphasized agriculture (1.0), light industry (2.0), and advanced industry (3.0). It is hoped that the model will enable Thailand to come out of the \u201cmiddle-income trap\u201d.<\/p>\n

In a press release<\/a> by the Board of Investment, AI can spur industry transformation.\u00a0According to Dr. Siridej Boonsaeng, Dean of the College of Advanced Manufacturing Innovation at King Mongkut\u2019s Institute of Technology Ladkrabang, \u201cThe future of production will be mass customization where data-driven AI and adaptable factory is the answer for survival\u201d.<\/p>\n

However, there are potential roadblocks to a successful AI implementation in Thailand. The country ranked<\/a> 56 out of 194 countries in Oxford Insight\u2019s Government AI Readiness Index in 2019. Cautionary measures need to be taken on behalf of businesses and the following presents some of the considerations:<\/p>\n