Turn taking and chatbots

Turn taking is a natural part of conversation that we subconsciously engage in order for the discourse to flow. Here is an example:

A: “Good morning”

B: “Morning. How are you? Good weekend?”

A: “Yes thanks, and you? How was Brighton?”

For the Cambridge main suite speaking exams, candidates are assessed on their turn-taking ability under the criteria of ‘Interactive Communication’. In other words, this means the candidates’ ability to:

  • Interact with the other candidate easily and effectively.
  • Listening to the other candidate and answering in a way that makes sense.
  • The ability to start a discussion and keep it going with their partner/s.
  • The ability to think of new ideas to add to the discussion.

Along with the onslaught of technological advances came advance in automated responses from portable digital devices. These conversational agents or dialogue systems are capable of single interactions or up to 6 task-oriented turns. An example of these dialogue agents would be Siri, and an example of a talk-oriented interaction would be: “Siri call Dad”.

Chatbots are not a ‘new’ invention per say. Eliza, created between 1964-1966 at the MIT was a natural language processing computer programme that demonstrated the same characteristics of chatbots day, but on a less sophisticated scale, and with less complex interaction. The aim of chatbot builders is to create natural language processing programmes that replicate human-human interaction by enabling more turns and therefore extended conversations.

The interesting challenge then becomes, how to use each turn as a springboard for the next, and ensure that each one prompts a response that has been pre-programmed, in order not to receive a generic message like: “I’m sorry, but I’m not sure what you mean by that”, when the user is expressing a specific request or a expressing a turn that is not recognised. More about chatbots soon!

IM: words in the air or recorded forever?

IM: words in the air or recorded forever? How many times have you misread a text or IM, or been misinterpreted yourself? Written spoken discourse leaves itself open to misinterpretation because the suprasegmental features are not apparent and neither is body language.

When we speak face-to-face with others, we are careful about what we say for fear of misinterpretation or offending the person/people you are conversing with. We therefore carefully and diplomatically communicate our message and use body language and features of connected speak to express ourselves.

Ironically though when we message others using one of the plethora of online messaging apps and services, or a mobile phone service providers’ texting service, we often send the message before we have had time to re-read it. It seems to be the case that the immediacy that instant messaging has brought the global messaging community has affected the way we communicate.

The written word is recorded, and the spoken is ephemeral discourse in the air, yet we pay attention to the ephemeral and not the recorded!

Just what do we expect from Chatbots?

Chatbots are the future of conversation intelligence, and can be used to stimulate human conversations. But just what do we expect from chatbots? On the one hand are those that firmly believe intelligent systems will dissipate the element of human interaction in years to come. On the other hand others revel in the delights of giving Siri instructions to challenge her intelligence and gauge the level of response.

Personally I feel that benefits for intelligent systems (chatbots) outweigh the disadvantages, but I am convinced that the advantages will depend on our behaviour and receptiveness to accept their merits. AI cynics were delighted when Microsoft’s Tay was morphed to demonstrate bad behaviour. At last there was proof to substantiate the argument in favour of the severe dangers of AI.

Users of Alexa were slightly disturbed by her random outbursts of laughter, to the extent that her code was re-written to disable a reaction when requested to do so, and to avoid reactions to false positives that try to trick her. This all leads to the question of the levels of humanness we expect from intelligent systems and chatbots, or more to the point the level of humanness we, as ‘humans’, are comfortable with accepting from ‘machines’.

Reflective video

Reflective video is an element of reflective practice I have waxed lyrical about for a long time. The ability to see oneself and analyse how we come across when we communicate, and if we are capable of transmitting a clear and succinct message that learners and/or attendees at a workshop/seminar can comprehend.

I was asked to make some ‘short’ videos about the future of teaching and learning, if technology influences and shapes how we communicate, and therefore the teaching and learning of languages.

Here is a trial run of one of the videos – it was supposed to be 2.5 – 3 minutes long. Reflective practice note to self – get to the point and make your message clear.

AI: a new currency or the next industrial revolution?

A question that has been on my lips recently is whether AI is set to be the next industrial revolution, or a new currency of the future.

AI Past

The industrial revolution as its name denotes, revolutionised modern industry and manufacturing as we know it today. When the internet emerged in the late 1980s it seemed unimaginable that less than 30 years later, wireless connections and digital devices would have such a pervasive presence in society. New inventions come and go, and technological innovations are created whether they are successful or not, but in most cases they are shaped by the demands of people.

The origins of AI date back to Turing’s computational machine more commonly known as the Turning machine, built in 1935, however the term was coined later in 1955 by McCarthy who defined it as “the science and engineering of making intelligent machines, especially intelligent computer programmes” (ibid 2001:02), in other words trying to understand human intelligence by using computers.

AI Present

During the last 80 years, advances in AI technology have reached astounding levels. It has clearly had a prolific impact on society, to the extent that it has been transformed into a tool in all aspects of life; from banking and email pop ups, to ‘personalised’ selected products, and Siri and Alexa the intelligent personal assistants, and chatbots.

AI Future

Both academic and business investigation and reporting in the field of AI, consider it to be one of the biggest influencers for the future of the market and society. Predicted revenues from AI are unprecedented, resulting in extensive funding and investment from private companies and governments, which highlights the significance of AI in society. China has recently announced they are building a $2.1 billion industrial park for AI research. The past year has witnessed an increasing amount of nations realising the importance of AI in shaping the economics of the future, some even consider it a currency. Bitcoins stand aside, AI is the new currency..



Reflections 2017

Reflections of 2017: The debate regarding the dangers of spending too many hours glued to an electronic device continue to bubble. The unknown abyss and potential of AI in its many guises continues to be explored. The fears of a robot-controlled world continue to rise. What will 2018 bring?!

Personally, I find all the above extremely exciting. Do I use my phone too much? I know I work too much, and because 80% of my work is online, I am obliged to use a digital device. This has become part of the natural shift in the plethora of work that has become created as geographical boarders are transcended by cyberspace and the power of technology, telecommunications, and IT. Just as technology is shaped by the society that uses it, tech very much shapes society and the way we interact and go about our day-to-day. I view technological developments as portals to opportunities that can be enhanced or were not previously available, especially in a teaching and learning context (whether that be English or dancing to Michael Jackson’s Thriller!!).

I will continue to explore how ed tech can support language learning this year, as I delve deeper into the AI and machine learning chasm. I will also wonder that if smoking hadn’t been banned in pubs and bars, if smartphones wouldn’t be the go to company we choose as we sit alone sipping a coffee contemplating the week, or waiting for a friend.

Learn to dance Thriller with NAO

Exams Catalunya 5th Annual ELT Conference – Reflections

Last weekend I attended the Exams Catalunya 5th Annual ELT Conference at ESADE in Barcelona. The theme of the conference was how to maximise interaction between learners and teachers to optimise learning outcomes. I was pleased to see that the role of Ed Tech as a tool to scaffold learning was given some attention. The audience was asked at the opening plenary to complete the following sentence, one which for me brings up many ideas:

Learning is ……………… with technology

Answers included, ‘fun’, ‘real’, ‘meaningful’, and ‘learner-centred’. Personally I think learning becomes attractive with technology, if, and that’s a large if, the learners use tech in their daily lives outside the classroom. I think this is an important factor to take into consideration when using educational technology for teaching and learning. Not all 50-year olds, and 6-year olds own their own, and/or feel confident with using technology, so its use needs to be specifically employed with clear pedagogical goals in mind.

One of the most interesting talks I attended, shared research about a project carried out with learners in a secondary school in Barcelona. The teacher was faced with a teaching puzzle where she had many proficient teenage learners, some with English-speaking parents, studying English in her Baccalaureate class, and was stuck for ideas about what to teach, that would include all levels. She took the brave decision to hand the teaching over to her students as a peer-teaching project, and the project was a huge success. Differentiation is an issue in many classrooms, and the obvious solution is to mix students with more and less proficient peers for different activities, so everybody learns from each other. This project took that idea further and demonstrated outstanding creativity from the learners.

The two main messages I took away from the conference were useful reminders to…

1: Never underestimate the creativity of the learners you have sat before you in the classroom.

2: Think carefully who your learners are and their background use with digital technology, before presuming they are happy and agile with its use.

Thank you to Exams Catalunya and all the presenters that gave the talks I attended.

Assessment for learning

Recent talks with colleagues working in the public education sector in the UK about SATs (Suite of Assessments), and my own experiences tutoring on pre-sessional courses, have given me a first-hand insight into the exhaustive measures some institutions employ to ‘promote’ learning through continuous formative assessment. The term they have coined is ‘assessment for learning’. Experience has demonstrated to me that the learning gains are limited compared with the time taken to prepare for the testing, the testing process, and of course the marking and feedback sessions.

A typical writing test could include learners being given several extracts from source texts to read and make notes on a week prior to the actual test. On test day, these notes are not permitted into the classroom and a new set of notes is given with a question to analyse. In my humble opinion, while learners will have read the texts and have a deeper understanding than seeing them for the first time, the test is in fact an evaluation of memory where they are desperately digging deep in their brains to retrieve the information about the points they deemed worthy of remembering. The question is analysed in groups, a draft plan is drawn up individually, and finally a 90-minute test is undertaken. Learners are notably exhausted after a testing process, which has essentially been drawn out over an entire week.

As an experiment, I tried an alternative approach where I gave learners 4 short extracts. In pairs each learner read 2 different texts and made notes. The notes were swapped with their partners who used them as a springboard to understand the 2 texts they did not read. Each learner proceeded to read the texts to accompany their partners’ notes to discover if they had identified all the key themes. A group discussion was held, a question was given, and learners wrote a short piece of discourse with a 40-minute time limit, to answer the question referring to the key themes identified previously, and citing as necessary. When the writing was completed, learners exchanged their scripts with a peer, and it was reviewed for content, accuracy of answering the question given, coherence, cohesion, stance, and argumentation. Another group discussion was held, and at this point I also participated with language support and academic guidance. This ‘think tank’ approach appeared to be effective and the feedback I received from the class this was tested with was positive. Including comments such as ”I learnt from my friends so it helped me feel confident to write”, “she was able to notice some additional points I didn’t see”

Nao (Softbank robotics) – Robots that write

After spending time with Nao (Softbank robotics) in February I am not in the slightest bit surprised at one of his many skills is the ability to write any word asked, and spell the word as he writes. Through speech recognition programming, the robot is able to perform many tasks, but the one of writing is a profound tool that can help those with literacy skill deficiencies, and of course those wanting to learn a language. Another interesting feature that will support my current research.