Search This Blog

Jan 27, 2023

Chatting with AI about AI

  • Me: "I am trying to find a metaphor that describes the feeling when you realize that most of your learning over many years is now becoming obsolete."
  • AI: "One possible metaphor for the feeling of realizing that one's hard-earned knowledge is becoming obsolete could be "watching a beautiful sandcastle wash away in the tide." This metaphor evokes the idea of something that was carefully built and treasured being swept away by forces beyond one's control, and also suggests a sense of nostalgia and loss."
OK, the metaphor is not that great. But we all should realize that language-generating AI will change our lives probably even more than Google and social media have changed it. The instant internet searches made access to knowledge radically democratic. Social media made production of content cheap and therefore very accessible. What AI is going to do is to make lower-level skills associated with formal writing, information search and summarizing, and language translation, obsolete. These are very hard news to swallow. We used to link lower-level and higher-level skills in one bundle. For centuries, the appropriate use of written language was an indicator of education and intelligence. We sort of knew that true intelligence, originality, and creativity were separate things. But in reality, and in most cases, if you were able to write grammatically correct, and learn the language conventions, you would be probably also good at more complex, more sophisticated tasks. Now this assumption goes out of the window. And it is not a pleasant discovery. OK I have a metaphor better than that AI provided: It is like watching hyper-inflation wiping out your life-time savings.

I know what is going to happen, and it is not good. Higher ed will start fighting another losing battle. We will look for algorithms to detect the use of AI. We will proctor exams, and make students write with pen and pencil. We will regress back to multiple choice testing, etc. This happens every time. As soon as the real-world changes and asks for something else, we keep teaching the old things, because that is what we have been doing before.

What should happen is something completely different. We need to reimagine how writing in real life will now happen with the help of AI. We need to explicitly teach our students how to use the AI in an ethical way, but also enjoy the potential for increased productivity. And finally, we should find forms of assessment that are not focused on originality of the text, but rather on originality of ideas. Some of the sacred cows, like learning APA format of references, is completely irrelevant now. The AI can fix your references in a second or flip them into another format. It is not a relevant skill anymore, like lighting fire with flintstone is not. Things like editing for grammar, summarizing the argument, or basic lit search are among those irrelevant skills. Therefore, we should spend more time on original, creative, and independent thinking. The question is, how do we do it?

I suspect it is still back to the Bloom's taxonomy. Evaluating and creating knowledge are still the higher goals. However, the definitions of evaluations and creating the new will change. New knowledge will not amount to new text. The quality of new knowledge will not be reduced to proper writing, or even include it.

No one is sure what is going to happen, but one thing is for certain. The world is changing.

So I asked the AI, if it thinks AI will change the world of work. Here is the response. It maybe a little biased, and terribly boring, but I must agree:

“AI has the potential to greatly impact the world of work. AI-powered automation, machine learning, and natural language processing are already being used in many industries to automate repetitive tasks, improve decision-making, and increase efficiency.

In the short-term, AI has the potential to automate many jobs, which could lead to job displacement. However, it also has the potential to create new jobs and make existing jobs more efficient and enjoyable.

In the long-term, AI has the potential to significantly change the nature of work, as well as the skills that are in demand. Companies will likely seek employees with skills in areas such as data analysis, machine learning, and software development to help them take advantage of AI technologies.

AI has the potential to improve productivity and create new opportunities for growth and innovation. However, it will also require a significant shift in the way we think about work and the skills that are needed to succeed in the future. It is important for governments, businesses, and educational institutions to work together to ensure that the benefits of AI are widely shared and that everyone has the opportunity to develop the skills they need to thrive in the new economy.”

Jan 23, 2023

Don’t assign homework to your dean or chair

Quite a few people enjoy assigning homework to their leaders - to program coordinators, chairs, deans, provosts, and presidents. At every meeting, at every opportunity, they will say: "We should do this or that," while clearly assuming that someone else has to implement it, somehow. And if the people in authority are hesitant or insufficiently enthusiastic, well, then it's clearly their incompetence or sabotage.

Well, this is not how higher ed, or any other organization, can work. To undertake a project of any significant size, someone must evaluate whether it is worth it (that is the ratio of effort over results), and if it aligns with strategic priorities. Someone must assess the scope of work and find resources – people mainly – to do it. Alternatively, the faculty member who has a brilliant idea can execute it on their own or lead a volunteer ad-hoc committee! Administrators and chairs cannot and should not prevent faculty from exercising their right to faculty governance and initiative. It is also totally fine to make a proposal to the entire group and ask academic leaders to support a project, including finding resources. Those kinds of proposals go through some sort of a debate, then evaluation, planning, etc. But it is critical for sustainable development not to chase every new idea that sounds good at the moment. That is the path to chaos and waste of resources. The entire concepts of mission and strategy are designed to prevent us from scattering attention and resources.

My advice to chairs and coordinators: recognize when people try to give you homework without a good reason. Do not promise anything you may regret. Gently push back by asking: OK, great, who is going to do this? My advice to faculty: keep generating ideas, but also invite others to critique and analyze them. Like a pair of jeans, a good idea has to be worn-down a little. Do not expect that someone else will pick up on your idea and just do it because you think it is great in general, regardless of context.

A fine line separates simple requests from assigning homework. However, here are some examples. I am happy to send you an existing document, but don’t ask me to compile a document especially for you just because you're curious. I am very happy to meet and discuss anything, but don’t ask me to prepare for the meeting in a particular way. I will answer any questions but won’t teach you how to do something complicated. I will match your idea with existing programs and resources, but will normally question the creation of any new program or resource. I will react quickly and use any opportunity to advance our mission but won’t do things that merely look good.

Little kids pass through the stage where they believe there are omnipotent adults who can do anything they want. And if they refuse to do something, it is because they are intentionally hostile towards them. Most of us outgrow this illusion and realize there is no omnipotent authority and unlimited resources at the disposal of those with authority. We live in a very imperfect world and work in very flawed organizations. A dose of realism makes for much better working relationships and fewer disappointments.

Jan 9, 2023

Weather and the human condition

The stormy, rainy days we've recently experienced have led me to contemplate how our planet's weather has shaped our evolution as a species. We learned to adapt to annual cycles, which are simply due to our planet's tilted axis. Furthermore, we developed resilience to the unpredictability of weather, knowing that any given day could be cold or warm, sunny or cloudy, calm or windy. One could easily imagine a different planet with no yearly patterns and a much more predictable weather system. Would such a planet cultivate a different kind of intelligence? However, it's our specific mix of predictable annual cycles and unpredictable daily variations that gave rise to our species. We mark our lives in years and take pleasure in being surprised by the shifting weather. Our civilization was largely shaped by our struggles to control these predictable and random changes. Our houses and clothes are just protective measures against adverse weather.

Perhaps the most profound consequence of our astronomical circumstances is this: we can predict with certainty the future weather for the next few months, but anything beyond a week or so remains unclear. For example, we know for sure that Sacramento will be hot by May, but no one can accurately predict the weather two weeks from now. Weather forecasts have slightly altered this situation, but not drastically. Our brains are pattern-seeking machines, yet they are consistently exposed to instances of pure randomness. The very need for free will may stem from weather unpredictability: when things are uncertain, we need to make decisions, which invariably involve assessing risks. It makes me wonder if intelligent life could exist on a planet without weather.

Weather teaches us to be adaptable, to change plans, and to always have a backup plan "in the event of inclement weather". The term "inclement" literally means "unmerciful". The unpredictability of atmospheric events planted a deep-seated belief in us that some divine force is behind a storm or a drought. However, it's more likely the other way around – weather has taught us that we can be unpredictable and random. The concept of free will likely mirrors the weather. We aspire to mirror the unpredictability of the skies. In many respects, we are merely reflections of our planet, despite our foolish belief in our own autonomy.