Hello,
Having worked with IT since the early 1980s, I recently tried my hand at some coding. Programming languages have changed and struggling with syntax and semantics of php, I decided to try Chat GPT to help me save time. Lo and behold, I was done in minutes! It even gave me fixes for some bugs with great explanations! This experience has led me to believe that AI systems need failproofing, else we are creating ‘covered pits’ into which we could figuratively fall. What does this mean?
Stretching it a bit, it means “No need to learn how to write code to solve problems. No need to know how to fix bugs, Large Language models will take care of that for me”. We are developing new skills at the cost of hollowing out some of our coding skills. Gen Z will start with ChatGPT or LLMs, not knowing or having worked with problem solving skills of coding from scratch as we had to in the past. The danger is that many of them will not even understand the way the AI generated code works. That is a major problem because it threatens traceability and explainability of code which is core to AI development today. Yet Gen Z needs to solve the problems of the future, which is exactly what the world needs.
The same scenario can be described for calculators which replaced slide rules and memorised multiplication tables in the end 70s. Are we better off today? Certainly, but only if technology and science progress ahead continuously providing us the fail proof technology foundation to work on.
When adequate investments are not available for technology, science innovation and technology support, technology begins to fail and the ‘covered pits’ start opening up. The recent outage of the Heathrow flight control systems in end Aug 2023 is a case in point. Reports indicate it was by caused by flight data which its system could not interpret1.The UK Treasury Committee launched its IT failures enquiry on 23 November 2018, following a series of high-profile service disruptions within the financial services sector. Most notably the TSB IT migration in 2018, which caused significant disruption to customers for a prolonged period of time2.
As transport infrastructure deteriorates, nations still chug along for many, many decades. Yet when nations with big tech companies start to de-grow, walls constructed between nations to limit trade by embargoes speed up that de-growth. Markets shrink for those very big tech companies on which technology progress depends. How will that affect the investments for technology and science growth of the big tech companies? What will be the priority for them, align with sales and the biggest markets, or with national policies driven by geo-political agenda?
Times have changed, world population is at over 8 billion, global warming is taking a heavy toll and age old rivalries are resulting in incessant wars. What the world could bear with a population of 2 billion in 1929, almost a century ago is quite different from what the world can bear today. In the 1929 -1933 Great Depression, the US economy shrank by 29%, unemployment rate reached 25%, consumer prices fell 25%, whole sale prices plummeted 32% and 7000 banks, nearly third of the US banking system collapsed3. HTTPS://Stlouisfed.org. Today, it could mean the perishing of millions across continents due to starvation, thirst and disease. Technology failure based catastrophes can result even if for a limited time. Extended catastrophes could mean threat to life at large scale. Let’s not add technology failure to the list.
So how can we cover those pits with strong concrete slabs? The three aspects to consider here are technology investments, fail proof systems and human competency levels. Companies themselves should realise the importance of backups, redundancy and fail proof systems for their technologies and invest in them adequately. Inclusion of these items in Risk management execution is important. Capital allocation balance between investments in sales, innovation and investment in ‘covering the pits’ should be supervised by the Boards of companies. There should be a fallback level of human technology knowledge and expertise available in companies as technology failure restoration insurance. National governments should also guide companies periodically in a collaborative fashion. The agenda for dialogue between politicians and CEOs should include ‘covering the pits’ as an item.
With the gigantic strides made by AI, the level of percolation of technology into our daily lives has multiplied manifold. Our dependence on technology is therefore at much higher risk. Let us work now to ensure that technology does not fail us before it becomes very late, like the state of global warming today.
References:
1. https://www.business-standard.com/amp/world-news/disruption-in-uk-air-traffic-control-caused-by-incorrect-flight-data-123083000515_1.html
2. UK treasury committee report on IT failures in the financial services sector 2019 to 2020.
3. How Bad Was the Great Depression? Gauging the Economic Impact – Federal Reserve Bank of St Louis HTTPS://Stlouisfed.org
© 2024
Most meaningful analysis . AI needs full focus of all concerned .