Hello All! This is the busiest time of the year for me ka!
วันอาทิตย์ที่ 28 กรกฎาคม พ.ศ. 2567
Busyness with Learning
วันเสาร์ที่ 13 กรกฎาคม พ.ศ. 2567
Learning more about AI (2)
Hello all. This week, I'll be real busy so let me share more about AI from an HBR article on ‘AI and Machine Learning: 4 Types of Gen AI Risk and How to Mitigate Them’ by Öykü Isik, Amit Joshi, and Lazaros Goutas, May 31, 2024, along with several others na ka.
According to the article, here are what we should stress:
1. Governments’ role: It points out that ‘governments are scrambling to come up with reasonable frameworks and laws to manage this technology and its downsides’ It proposes a high-level framework that will provide executives with a way of classifying the potential challenges within the gen AI landscape and then mitigating them.
This leads me to think of what Khun Tonson Santitarn Sathirathai said in several programs, suggesting our authorities not to have too rigid laws, which could hinder progress of our efforts to learn and excel in the use of AI ka.
2. Generative AI Risks can be classified based on intent (deliberate malpractices) and usage that are put into 4 types, namely, misuse (unethical or illegal exploitation of gen AI capabilities for harmful purposes), misapply (prioritizes plausibility over accuracy and can create inaccurate outputs – hallucination), misrepresent (output created by a third party is purposefully used and disseminated, despite questions about credibility or authenticity), misadventure (accidentally consumed and shared by users who are not aware of its inauthenticity).
3. The authors call for leaders in both public and private enterprises to become proactive and to mitigate these risks as follows:
3.1 Mitigating content creation risks to avoid the misuse and misapplication by developing the capabilities to detect, identify, and prevent the spread of such potentially misleading content.
Actions include strategies to align between organizational values and AI principles, mandate all entities that create gen AI content to watermark their gen AI output for transparency traceability and trust plus the empowerment of users to confidently make judgements on the authenticity of the content they come across, and create a controlled gen AI environment within the organization (curate training datasets, ensure their de-biasedness, and have privacy measures).
3.2 Mitigating content consumption risks to avoid misinterpretation and misrepresentation
Actions include gen AI demystification and awareness training opportunities, validate AI output through labeling and warning mechanisms, damage mitigation plan setups for situations that are not contained.
In the AI world today, it seems that the private sector has stepped up much faster than the governments, which is similar in our country….though we seriously need more ka….instill sense of urgency and understanding in our key leaders/authorities for proper policies and support, upskill our leaders and people for experiments and actions, invest in our infrastructure, and speed up in all dimensions.
วันอาทิตย์ที่ 7 กรกฎาคม พ.ศ. 2567
Learning more about AI (1)
Good morning ka. Following a request for my podcast program to share info and thoughts about AI in education, I started listening and reading a bit more and here’s one of the articles to share with you ka.
One is from Harvard Business Review on ‘AI and Machine Learning: 4 Types of Gen AI Risk and How to Mitigate Them’ by Öykü Isik, Amit Joshi, and Lazaros Goutas, May 31, 2024. Also watched some YouTube programs kaa!
Key points from the overall learning these days are:
1. The first sentence of the summary is what I’d like to quote loei….it reads, ‘Many organizations are understandably hesitant to adopt gen AI applications, citing concerns about privacy and security threats, copyright infringement, the possibility of bias and discrimination in its outputs, and other hazards.’
My own thinking is the word ‘hesitant’ as I feel we don’t really have enough knowledge and understanding. I forget where I listened it from but one podcaster was saying that leaders are slow and don’t really think of using it seriously….only instruct their teams to learn and get started which isn’t practical ka.
No matter what, leaders and their engagement are crucial, noting that ‘approximately 70% of respondents (from 2,500 executives by Gartner last spring) reported that their organizations are exploring how to integrate generative AI into their organizations while a published Stanford AI Index Report shows that the global AI adoption rates are higher in all surveyed regions.
Our public sector is far too slow and hopefully our private sector can take a strong lead though they obviously strive to have AI integration. Our government must have qualified experts around to help, with political will to come up with decent support and not too rigid regulations.
2. I listened to Khun Krating KBTG and liked his mindset, which is inspiring. He said this is the best time for most of us to work on AI-related development as it’s something quite new and most people are more or less at the same level of learning so we need to gain from this advantage.
I’d like to add, ‘Go for it fast so we won’t be left too far behind for others to look back and help ka!’
3. In addition to worries about learning, speed, and consequences of the use of AI, one more thing is put clearly in the HBR article…. ‘……downsides of widespread adoption of gen AI range from mildly annoying (such as personalized spam) to truly catastrophic (such as the fast depletion of water sources from vulnerable parts of the planet to support the largest data centers ever constructed)!
How to balance the benefits and the damages is one very important issue to manage with care!
Going for a bit too long and I'll continue the second part soon ka.
วันพฤหัสบดีที่ 4 กรกฎาคม พ.ศ. 2567
IAU's Global Survey on IZN
Hello everyone! Busy busy time for me as I’m helping organize a program for 16 Fulbright-Hays high school teachers who’ll be visiting Thailand for 3 weeks ka.
In any case, thanks to Ajarn Piniti, former SG of Higher Education Commission for sending to me the 6th IAU Global Survey Report on Internationalization of Higher Education: Current Trends and Future Scenarios by Giorgio Marinoni and Siro Bartolome Pina Cardona, 2024 ka.
I’ve read the summary and conclusion only to come up with the following:
- Out of the 722 HEIs from 110 countries and territories surveyed, I’m kind of disappointed to see that our Asia & Pacific higher education institutions are underrepresented.
I feel this could mean a lower priority on IZN as I haven’t heard universities discuss this topic as much as before ka.
- The survey shows the widespread nature of IZN as a strategic process….it’s time kaaa!
- It also reveals that the main IZN drivers are institutional leadership and the international office though unfortunately in Thailand, I think our international offices aren’t regarded that important and serious reviews about the proper role of international offices must be made to gain the real benefits of IZN ka.
- “Enhanced international cooperation and capacity building” remains the most important, followed by “Increased global, international and intercultural knowledge, skills and competences for both students and staff” as IZN benefits globally in all regions except North America.
My concern is that we’ve seen too many MOU signing ceremonies again ….so scary even after having straightened understanding of the importance of MOUs three decades ago kaa!
- In terms of the risks of IZN, they include Increased workload for academic and administrative staff.
I think this needs understanding of senior leaders for their necessary support and involvement as well ka.
- Some positive news ka:
- 77% of HEIs have elaborated an IZN strategy, Europe as the highest relating to policy/strategy while most have revised or issued or been undergoing revision of their IZN policy/strategy.
- 74% of HEIs have a monitoring and evaluation framework to assess progress.
- 75% of respondents acknowledged a noticeable increase in the importance of internationalizing the curriculum at home within their institution over the past five years.
- “International research collaboration and outputs (e.g. international co-publications)”, “Outgoing mobility opportunities/learning experiences for students (study abroad, international internships and placements, etc.)”, and “International development and capacity building projects” have grown increasingly important.
- In North America the majority of HEIs indicated that the increase in online collaboration has led to the inclusion of a new online component to existing joint degree programs with international partners.
- The regional results confirm the overall positive IZN impact on promoting intercultural understanding and reducing racism and xenophobia in all regions of the world.
- Asia & Pacific is clearly the region where the link between IZN and sustainable development is more advanced, as 52% of institutions in that region indicated that they have a policy or strategy in place to use IZN as a means to support sustainable development.
Some good progress on IZN and hope there'll be more development plus related info in our country and region in the next survey ka.