AI snake oil in education

Pattern recognition is the holy grail of AI and machine learning. As such, its practical implementation is being explored everywhere- education, healthcare and “self-driving cars”. It is a solution looking for a problem.

One of the reasons why the vacuous debates happen around the AI is because of its liberal constructs. Many startups have found the perfect template to perpetuate these issues at a mass scale. Copy the idea; get funding and hire the PR firms that help to supplant it into “opinion columns”. Here are two write-ups that appeared in Financial Times, and I am quoting them selectively. Notice the choice of words (emphasised for clarity).

The first one:

The world of education is heading for an unpredictable destination.New technologies, especially pattern-recognition software commonly referred to as artificial intelligence, are descending upon the learning process, changing how we retain and test knowledge.

Artificial intelligence in education relies on the idea that software can recognise patterns in students’ performance, highlighting strengths and weaknesses while enabling them to improve more rapidly.

These are “forward-looking statements” because they have not been formally tested. The outcomes of education (didactic methodology), for example, are related to getting a “job” and not “learning”. As these abstract measures are hard to quantify, it is a relatively grey area and hence “ripe for disruption”.

Here’s something more alarming:

Up Learn’s software provides animated video lessons, which continuously test students, and it guarantees an A*/A equivalent exam grade — or your money back.

Is the job of FT to inform or do this subtle advertising? How much do they charge to weave in the narrative around PR handouts?

In developing economies such as India, shortages of teachers are often cited in support of introducing technological solutions. However, there are concerns over the use of technological approaches in developing economies, especially if it is driven by profit-seeking organisations and is perceived to reduce contact with teachers.

This statement is patently wrong. In India, the teacher’s appointments are political. The political masters dilly dally because they are primarily bankrupt. AI won’t replace humans but will not supplement them either.

As such, the rise of the “ethicists” in the Universities that are cashing on the trend.

The second one:

A quarter of teachers in England work 60 hours a week, according to a recent University College London report. The overall average was 47 hours during term-time, which is eight hours more than comparable OECD countries.

While 98 per cent of teachers say they enjoy teaching, they spend less than half their time in the classroom because of the burdens of administrative work, planning lessons and marking, according to a report by Ofsted, the government’s inspectorate.

This is what is called as an attempt to create a “selective outrage”- trying to influence teachers that they are “overworked” and cherry-picking the “facts” from a government report.

“Tech . . . can empower the teacher to spend more time teaching, freeing them up to focus on skills like critical thinking and collaboration,” says Priya Lakhani, chief executive of Century Tech, a learning platform for schools and universities.

Do you spot patterns now?

That’s why you need an extremely healthy dose of scepticism to deal with these “reports”. This is yellow journalism- PR narratives woven as an opinion without the declaration of the conflicts of interest.

Last but not least is the story from India on how an ed-tech platform has hoodwinked the parents into buying its subscription. (I apologise that it is behind a paywall).

In the last half-decade, Byju’s has revolutionized learning, which remains the heart of its operations. But sales is the oxygen it breathes. It might be India’s most valuable education technology company, but start picking it apart, and you will find the skeleton of a hyper-competitive startup, driven by sales numbers, buried deep inside.

Here’s another interesting blurb:

One of the most popular questions BDEs ask children in grades 6 and above is about the number of points on a circle.

“Most children fail to answer that question,” says one of the BDEs. “Then, we explain that a circle has unlimited points.” 

All the valuations for the company are just paper valuations. The product is the subscriptions and fails to honour the complaints on the ground. The article went on to explain the sales process, but its AI snake oil is worse. It was covered in detail elsewhere with an in-depth interview with sectoral experts debunking its myth of “achievement of learning objectives”.

There’s something similar in healthcare, too.