Posted by After three years of thinking about it, and three months of planning, I have finally begun upgrading my human-computer interaction course, over on Udemy, one baby-lecture at a time.
Creating a baby-lecture for me, is like the puzzle of the chicken and egg, and which comes first. Was it the chicken or the egg? The script or the video?
I know that I need a script to make a video, as otherwise I can lose hours of time editing, but then I need a video in order to generate a script otherwise I can lose hours trying to think up what to say. I compromise and spend a lot of my time sitting in front of my camera opening and closing my mouth whilst I figure out the best way to say things, as I demonstrate in the 12 seconds of video below:
It takes me one minute to speak 123 words. I know the recommended average is 150 words per minute but having a glorious northern accent, especially when I get excited, it can sometimes be hard for people who do not hail from the north of England, to follow what I say, so I try to slow it down and make every word count.
I put that little video on YouTube as a short and it got 1.9k hits overnight, funnily enough all my longer videos about HCI didn’t get that many hits not even a gacha one about my CV. So, it just goes to show, I have no idea what is popular, but apparently the Udemy advice is to market yourself over there to drive students to your course on Udemy, as Udemy is just a marketplace and I have to shout: Roll up roll up. The thing is though, human-computer interaction (HCI) is not for everyone so I can’t imagine it doing well on YouTube. I mean I don’t even watch HCI videos over there. I subscribe to channels about Tarot, meditation, and a couple of ‘get fit in two minutes a day’ channels. Though, I roared with laughter when @Ofalafel added fart noises to some fitness videos on Instagram. Oh and I do like watching very long lectures on AI and robotics given by academics in an academic setting on YouTube, not programmes by the BBC etc., who would sex it up and simplify it for entertainment purposes.
The funniest comment I got on YouTube was: Are you AI?
What am I trying to say?
It takes me the same amount of time to create a three-minute video as it would do to create a whole long blog which I would post on here. If I am really going for it, it takes a week.
In either case I am answering the question: What am I trying to say? Sometimes I have no idea what I want to say until I have said it, and other times I don’t know how to begin, until I have begun. The same goes with blogging really, as I often have the vaguest idea of what I want to write about and don’t know until the end what it is I have said. And, even then I may think to myself: Oh that’s not what I wanted to say.
Last month, I tried to microblog on Twitter instead of writing one blog here just to try something new. I tweeted out my ideas one tweet one point at a time as if I was writing a blog or creating a video. The subject was about the length of a tweet, the length of a sentence, versus the length of a blog post or 4000 characters which is the maximum tweet size that the new micro-blogging facility Twitter is offering to blue-tick people.
I found it hard to micro-blog, or think in snippets, because I had more than one idea to express so it was rather like writing a list, but it does work better than reading a great big ‘micro blog’ of 4000 characters ~1000 words in the middle of a stream of Tweets, I would probably skip past it if it was in video (~ 5 minutes) or words. That said, I enjoyed the experiment even if I came to a different conclusion to the one I set off thinking I would arrive at:
But then, that is not new whether videoing or blogging or even chatting. I often end up with a: Mmm that’s not what I wanted to say. But after blogging about it, the conclusion seems inevitable though unexpected. Quite the dichotomy. The whole experiment confirmed what I was thinking, I prefer my lectures to be under five minutes for my Udemy upgrade, or human-computer interaction 2.0 as I like to call it.
I was thinking to document the journey too this time as I would have liked to do that way back last time, but by the time I had wrangled my ideas into a form which I liked, I was too tired to do the journey bit and just wrote a couple of blogs instead. This time I wouldn’t mind doing some videos about the videos.
Aha also, it’s another chicken and egg, I find it much easier to improve on something which already exists than to create something brand new. But creating something brand new allows me to improve on something which already exists. It’s all very yin and yang, backwards and forwards, like a dance, and I like it.
Last time, I made so many decisions and dropped so many ideas that it took a while to remember them all (as I always fling all my notes out afterwards but regretted it this time) when I began wondering how best to upgrade the course in December 2022. The minute I finished uploading the course back in 2020 I had so many ideas of how to improve it but I had to wait as I moved house, and felt that I should learn to plaster and do DIY to fix things up in my real life, before I settled back down to doing my course in my virtual one.
I also spent some time on the Udemy forums to see how the most successful people do their course upgrades, and read that one course developer said that for them to develop a five-hour course it takes 193 hours, another one said that it takes them 15 hours to produce one hour of content.
This is reassuring as last time, it took me five months to produce a course, though it was lockdown and I was completely distracted figuring out school work for the home schooling, and spending time in my ‘office’ which turned out to be a rug on the trampoline under the tree in our back garden, and then sorting and packing up the flat to move. In total my course took five months for just under two hours, and all that time I was berating myself for not going faster. Though why I should have gone faster I don’t know it’s not like anyone was waiting for it.
I remember feeling demotivated too because I read on the Udemy website that it takes three weeks to create a course, which seems like no time at all, even say working full time on it, say 40 hours a week, that is only 120 hours. I literally spent that just thinking about creating a course. I have sometimes spent 120 hours thinking about a blog I want to write.
AI at the WI
The other week, I gave a talk about artificial intelligence (AI) to my lovely local Women’s Institute (WI), of which I am a joyful member. I want to write a blog about it at some point as it was a wonderful evening, though I can only speak for myself.
My talk came in just under 70 minutes and I loved thinking about it, developing it and delivering it but since I was trying to condense something I may have delivered in a university over 12 weeks into one hour, it took me a long time to create. Thinking about it now, I did it over three weeks, and even if I don’t add in those times I was lying in bed in the middle of the night – I don’t sleep great nowadays – thinking about it, I could easily count up to 120 hours. But I have loved AI since way back when, and so it was no hardship to think: Oh I can say this, ooooh and I can say that.
Though my husband got a bit fed up and said to our girls one day: Make it stop. As I was sitting there at dinner talking about Marvin Minsky:
Did you know he believed that it was possible that we were all living in a computer simulation?
Apparently, Elon Musk feels the same way.
And, do you know that Musk’s companies had received over $4.9 billion by 2015 from the US government?
A government who cannot invest any money into a health care system for its citizens but can blow billions without any reason on AI!
And, what a lovely time I had delivering the actual talk to people who wanted to hear it (thank you, wonderful WI) – to be fair though to my husband, the WI heard it just the once, and my hubby heard it a lot and knows some of it already as he did AI. (We met when I was working at the AI Lab, EPFL, and he was visiting a colleague of mine, my hubby’s friend with whom he had studied AI in Edinburgh). My daughters too are now experts in AI which is amazing as they are listening though they often don’t look like it, and they have amazing minds like sponges absorbing all that information.
And if Marvin Minsky is right, perhaps I am AI? I could reply to that comment on YouTube.
The following day after I gave my talk, I was a bit sad it was over until I realised that I could turn it into a whole section for the HCI course, as AI is the reason I got started in HCI in the first place, and voila, another six baby-lectures which knowing me could turn into another month’s work. Or hopefully not as I have all the materials, but we will see as for me, making a baby lecture expands to fit the time allocated and that includes swizzling on the chair and talking to the cats.
Eighteen minutes and counting
So far, I have created seven new lectures, only five of which I have decided to use and upload, which adds up to (drumroll) a sum total of 18 minutes, and since I have another 35 baby lectures planned, without counting the AI lot, that is a ways to go yet, hobbiteses. By my calculations, 35/5 = 7 and 7 * 18 = 126 minutes which is two hours and six minutes might take me anywhere from 30 hours to 193 hours (depending on the Udemy expert: one hour = 15 hours prep or five hours = 193 hours prep) and since my 30 hours is never a whole week as I have other commitments then 30 hours is not a week, it could a month.
Regardless of the time, I am loving every minute, even on those days when can’t think of anything to say, as for me, right now showing up is enough, since anything might happen and often does.
Some course developers develop their whole new range of lectures and then upload them, I decided to do mine section by section and keep my students informed along the way. I like this as I have accountability and the idea that someone is waiting on me, regardless if that is true or not. Originally, I did think to do all the lectures and then upload them, but I never got started, I like external deadlines even if they are artificial and of my own making. So it would seem that I am baby-steps all the way to my upgraded course.
Next week I will start the I/O section and am looking forward to daydreaming about it over the weekend.
Come think about HCI with me for free, click on this link: https://www.udemy.com/course/human-computer-interactions/?couponCode=YOUTUBE (expires 4/4/23) and watch the course grow one baby-lecture at a time.
