I have created a link to a Prezi on my explorations in algorithms, I initially explored IBM Watson chat bots and Microsofts own azure version, but they are not as off the shelf as promised.
Brilliant artefact, Jon! Such insightful thoughts to frame your explorations around – ‘the idea of an algorithm as a cultural presence’ (Beer 2017) and the notion of ‘the algorithm as embodied’. Also, some interesting and, at times, amusing results! Perhaps reflecting the current polarisation of views we so often hear about (e.g. ‘Boris Johnson is a toe rag/genius’ being two of those results)!
I did a very small amount of playing with the Google search autocomplete with ‘is edtech…’, and it seemed to largely promote a profit-driven view, rather than a particular focus on education. Brings to mind how sometimes certain views or biases can be reinforced, having looked at Noble’s (2018) ‘Algorithms of Oppression’: https://nyupress.org/9781479837243/algorithms-of-oppression/
With regards to Spotify, that’s a fascinating insight into how they quantify music around terms like “speechiness” and “valence”…not terms I’d normally think of! I often find my musical tastes are quite unpredictable, while Spotify tries to play “more of the same”. I wonder if this is accounted for in the algorithm, and would be the same for one that picked academic papers for us to read?
That’s a fascinating piece of work, Jon! The auto complete is fun, many comedians build their shows around this algorithm. It was very wise of you to search both for a kind of neutral topic and a more politically coloured. I didn’t expect the results to be so ‘diplomatic’ either. They must be location-specific too. I tried the same collocations from here and there were zero matches.
You intrigued me by the fact that Spotify can ‘read you’. I feel I need to try it. Your idea to relate a similar algorithm to education is really trendy.This is exactly what recommender systems within LMSs or MOOC platforms try to do, not always successfully though. Since machines are insensitive to the context and human intent, I think it’ is still a long way to go until they’ll always ‘get it right for us’.
Thanks for sharing your discoveries with us, very insightful!
Crouchipuss
I love that the whole Carrot 100 thing comes up for Boris Johnson. That’s a game to play with Alexa’s translation service. “Alexa, translate Boris Johnson Carrot One Hundred into Welsh”… The results are quite similar to the thing you mention about Gordon Brown 🙂
Thanks Jon, very interesting artefact and the conclusion for application in education is compelling. That approach would definitely support the push for personalised learning. I wonder how much further personalisation has to go and how it will change public education? The old school might argue that reading something you don’t like or aren’t interested in is an essential part of education. If an algorithm removes those readings from a students reading list based upon what they have disliked reading before, what unforeseen results might this have?
Brilliant artefact, Jon! Such insightful thoughts to frame your explorations around – ‘the idea of an algorithm as a cultural presence’ (Beer 2017) and the notion of ‘the algorithm as embodied’. Also, some interesting and, at times, amusing results! Perhaps reflecting the current polarisation of views we so often hear about (e.g. ‘Boris Johnson is a toe rag/genius’ being two of those results)!
I did a very small amount of playing with the Google search autocomplete with ‘is edtech…’, and it seemed to largely promote a profit-driven view, rather than a particular focus on education. Brings to mind how sometimes certain views or biases can be reinforced, having looked at Noble’s (2018) ‘Algorithms of Oppression’:
https://nyupress.org/9781479837243/algorithms-of-oppression/
With regards to Spotify, that’s a fascinating insight into how they quantify music around terms like “speechiness” and “valence”…not terms I’d normally think of! I often find my musical tastes are quite unpredictable, while Spotify tries to play “more of the same”. I wonder if this is accounted for in the algorithm, and would be the same for one that picked academic papers for us to read?
Insightful thoughts and detailed explorations presented really clearly – thank you!
That’s a fascinating piece of work, Jon! The auto complete is fun, many comedians build their shows around this algorithm. It was very wise of you to search both for a kind of neutral topic and a more politically coloured. I didn’t expect the results to be so ‘diplomatic’ either. They must be location-specific too. I tried the same collocations from here and there were zero matches.
You intrigued me by the fact that Spotify can ‘read you’. I feel I need to try it. Your idea to relate a similar algorithm to education is really trendy.This is exactly what recommender systems within LMSs or MOOC platforms try to do, not always successfully though. Since machines are insensitive to the context and human intent, I think it’ is still a long way to go until they’ll always ‘get it right for us’.
Thanks for sharing your discoveries with us, very insightful!
I love that the whole Carrot 100 thing comes up for Boris Johnson. That’s a game to play with Alexa’s translation service. “Alexa, translate Boris Johnson Carrot One Hundred into Welsh”… The results are quite similar to the thing you mention about Gordon Brown 🙂
Pingback: Week 9 – Summary – Jon Jack’s EDC lifestream
Thanks Jon, very interesting artefact and the conclusion for application in education is compelling. That approach would definitely support the push for personalised learning. I wonder how much further personalisation has to go and how it will change public education? The old school might argue that reading something you don’t like or aren’t interested in is an essential part of education. If an algorithm removes those readings from a students reading list based upon what they have disliked reading before, what unforeseen results might this have?
If you ever want to revisit the chatbot stuff, there has been some progress in that area with QBot in MS Teams :
https://youtu.be/NcbQ2UK69Tc
and:
https://github.com/unsw-edu-au/QBot