Students’ self-awareness is on the rise – that’s my takeaway from the teaching of Finance, Data Science and AI development at the UK university, at both undergraduate and postgraduate levels, over the past three years.

This is a strong trend among the younger generation of future economists and financiers. They are becoming deeply responsible for their professional background. 

Today’s university students are more eager than ever to acquire cutting-edge knowledge from a multidisciplinary perspective, even when it requires intense study and rapid immersion into unchartered territories of expert skills and STEM (Science, Technology, Engineering, and Mathematics) proficiency.

I see today’s students showing greater curiosity about complex problem-solving grounded in interdisciplinary empirical approaches and a strong theoretical foundation. 

That is why students are increasingly reluctant to engage with outdated or irrelevant material and study cases. And we - university teachers - should do our best to earn their respect through our professionalism and expertise.

After all, our students are the primary beneficiaries and stakeholders who shape the climate of the entire university. As students grow more knowledgeable by the day, academic educators should meet the strong demand for up-to-date expertise. Fortunately, our students remain a source of hope for a better world despite the challenges we face today.

Published on Substack.com  "TradFi and DeFi: Broaden Horizons" (Note 151) by Dr. Olga D. Khon

With the rise of artificial intelligence (AI), the number of amateurs sliding on a puffed-off AI hype is booming. In short, the public is often misled or deceived by overuse of the “AI” terminology.  That’s a substitution of concepts – in more than 90% of cases when someone refers to the buzzword “AI”, they actually mean chatbots such as ChatGPT, Copilot, Claude, Perplexity or Gemini. In other words, human interaction with chatbots through conversational interface. So that an immense AI field is being mistakenly substituted by the narrower area of Generative AI (Gen AI) and Agentic AI.

To recall, both Gen AI and Agentic AI are based on large language models (LLMs) technology - an advancement in Deep Learning. This technology allows non-technical users to make requests in natural, human-like language. As a result, anyone can interact with chatbot-assistant and obtain a wide range of generated outputs, from texts, advanced analytics and modelling to images, videos and soundtracks. 

Therefore, for our economics and finance graduates, the use of LLM-chatbots is unlikely to become a strong competitive advantage. If anyone can use pre-trained LLMs to produce similar outputs, then LLM-chatbot functionality effectively equalizes all users (in some cases, even globally). From employers’ perspective, if a candidate’s only skill is prompting LLM-chatbots - regardless of context - the rationale for hiring them becomes rather questionable.

However, the greatest shock comes when AI does not make us smarter - as it is often assumed to do - but instead produces the opposite effect. These are the situations where the rush toward AI hype leads to the incorporation of Gen AI tools almost everywhere, often unnecessarily.  At universities, such an approach diminishes students' analytical and technical skills, shifting their mental effort - let alone creative thinking- toward chatbots. The habit risks becoming destructive for future generations. It also threatens sustainability outcomes in higher education due to overuse of scarce resources (e.g., energy and clean water). Finally, it raises questions about the relevance of higher education in a rapidly transforming job market that demands the more advanced analytical and technical skills from economics and finance graduates - skills that go far beyond simply operating chatbots.

Therefore, students nowadays should be extremely cautious while choosing the university educational programs. Even if there is a promise to teach you AI, it’s better to double check the components of that AI expertise beforehand. And certainly, the trivial chatbot-prompting isn't worth either your time or tuition fees in an economics and finance degree.

Published on Substack.com  "TradFi and DeFi: Broaden Horizons" (Note 150) by Dr. Olga D. Khon

No doubt the bar for graduate economists and financiers has been raised to advanced 𝗗𝗮𝘁𝗮 𝗔𝗻𝗮𝗹𝘆𝘁𝗶𝗰𝘀 and 𝗗𝗮𝘁𝗮 𝗦𝗰𝗶𝗲𝗻𝗰𝗲. It already includes popular programming languages, like 𝗣𝘆𝘁𝗵𝗼𝗻, 𝗦𝗤𝗟 𝗮𝗻𝗱 𝗥, but on top of it is𝗺𝗮𝗰𝗵𝗶𝗻𝗲 𝗹𝗲𝗮𝗿𝗻𝗶𝗻𝗴 as the backbone of 𝗮𝗿𝘁𝗶𝗳𝗶𝗰𝗶𝗮𝗹 𝗶𝗻𝘁𝗲𝗹𝗹𝗶𝗴𝗲𝗻𝗰𝗲 (𝗔𝗜) expertise. That’s the common requirement for economists and financiers to land a job nowadays.

Based on my experience delivering courses in machine learning and Data Science for economics and finance students, the core of successful teaching lies at the very beginning of the student journey in programming and data analytics. Student training should guarantee a stress-free entrance to the uncharted territory of programming languages first. Then, advanced statistics and econometrics, along with data analytics concepts. And, finally, machine learning and deep learning – to unveil the multi-layered AI paradigm (beyond chatbot interaction through human-like language only).

Otherwise, students may be discouraged from making any progress in the entire field of Data Analytics, Data Science, and AI Development. So that, sometimes decent students, who truly deserve and could succeed in the discipline, could never enter our classroom, being frightened by previous negative experience or overloading information available online.

Simply put, tutors should ensure students understand how to and why to code at all at every instance (within each line of code). It should go smoothly as any language teaching, with careful calibration from day one, to harvest student interest and learning excitement in their step-by-step progress toward the professional qualification in this multidisciplinary field of economics and finance, AI and Data Science.

Published on Substack.com  "TradFi and DeFi: Broaden Horizons" (Note 149) by Dr. Olga D. Khon

Artificial intelligence (AI), particularly generative AI (GenAI), is causing significant disruption in job markets, and higher education is no exception. To meet this challenge, academic circles have developed two extreme points of view on AI.

One assumes it can exist in a kind of parallel universe - remaining largely unchanged while still coexisting with innovation. Although it is clearly a dead end, this position still persists in many universities and across numerous educational modules today.

The other position - also a rather superficial one - is the idea of replacing every decisionmaker and managerial leader across all possible fields with people who hold computer science degrees. Unfortunately, this extreme view is also destined to fail.

First, a computer science degree does not necessarily guarantee strong qualifications in AI, since true expertise requires deep knowledge of both AI algorithms and computer programming.

Second, mechanically substituting domain knowledge professionals with general programmers would rapidly erode the quality of skills and expertise in those fields. Imagine replacing an art professional - a musician, a film director, or even an economist - with a computer programmer who barely understands the domain. The result would inevitably be disastrous for the field and for the development of qualified personnel.

Besides, as an economist and data scientist, I have witnessed costly errors and financial misalignments caused by IT specialists in several blockchain companies. It is comparable to putting someone with a pedagogy degree in charge of an academic department of finance. With no offense to any professional field, it is primarily incompetence that raises serious concerns.

Finally, three years after the first release of ChatGPT, the rise of Artificial Ignorance (including AI hallucinations and AI sycophancy) has made it clear to most people that relying solely on large language model (LLM) chatbot communication is no longer a viable option.

The unspoken problem is that highly educated computer scientists tend to remain within the AI industry for various reasons - from better career opportunities to more advanced professional development. So when you see someone with a computer science degree seeking to lead, say, a department of art or economics, the first concern that arises is their actual qualification. Are they truly competent in computer science, or do they belong to a segment of poorly qualified personnel who struggled to find a position in their own field and now exploit the AI hype while undermining colleagues in other industries?

Hence, the solution is, as always, interdisciplinary and grounded in essential qualifications. Consider Arts, Film Direction, Design, and other Creative Industries, or Economics and Finance - these disciplines should be led by professionals with deep domain expertise who either expand their skills in AI (beyond simple chatbot use) or have the vision to collaborate with top AI specialists in a genuine symbiosis of multidisciplinary knowledge.

We should not forget: domain knowledge comes first, and applied AI follows.

Published on Medium.com  "TradFi and DeFi: Broaden Horizons" (Note 148) by Dr. Olga D. Khon