It's wild that the Bloom study about 1:1 tutoring has been around for over 40 years, teachers have been advocating for increased staff & small class sizes the entire time, yet it's re-emerged as the TechBro mantra to answer education inequality & "democratization" with AI tutors. 😵‍💫 - ThreadSky

covingtonedu.bsky.social • 35 days ago

It's wild that the Bloom study about 1:1 tutoring has been around for over 40 years, teachers have been advocating for increased staff & small class sizes the entire time, yet it's re-emerged as the TechBro mantra to answer education inequality & "democratization" with AI tutors. 😵‍💫

Comments

covingtonedu.bsky.social•35 days ago

The cause of this rant on a frigid January Sunday morning: https://www.psychologytoday.com/intl/blog/the-digital-self/202501/can-ai-condense-two-years-of-learning-into-six-weeks

covingtonedu.bsky.social•35 days ago

The author, John Nosta, is the founder of NostaLab, a tech firm that, from what I can tell, is involved in every Silicon Valley pseudoscience imaginable.🙄

covingtonedu.bsky.social•35 days ago

Famously, "democratization" is when a single guy owns a proprietary technology that doesn't require paying human beings for their labor.

covingtonedu.bsky.social•35 days ago

But *why* are there such socioeconomic gaps in Lagos, Appalachia, and El Salvador, John? Whyyyy?

Turns out there are dump trucks full of money to be made in creating problems AND selling solutions.

elizabethwithaz.bsky.social•35 days ago

They should fix the respecting cultural and contextual nuances thing before bringing education into it.

breweredu.bsky.social•35 days ago

Creating problems and selling solutions is the critique I needed this morning.

Great thread!

bmorebaker42.bsky.social•35 days ago

The people that want AI teachers would never send their kids to an AI teacher.

bthily.bsky.social•35 days ago

Good to see some LinkedIn stuff being debunked here.

overit22.bsky.social•35 days ago

How many buzz words can you string together in one article?!? I did find it funny that they conceded that “Teachers mitigated challenges like misinformation by emphasizing responsible AI usage and critical thinking.” How exactly do you roll out this miracle cure when it’s ripe for abuse? Oh wait…

stephstephking.bsky.social•35 days ago

I clicked all the way through to the study and don't see the data to support their claim at all? It looks like the treatment group did mildly better than the control, but certainly nothing like "two years in six weeks"

covingtonedu.bsky.social•35 days ago

This is the standard deviation chart that the linked World Bank blog includes, and since they seem to lump English, "AI skills", and "digital skills" together it seems super unhelpful?

Like, what baseline for "AI skills" did Nigerian children have beforehand?

https://blogs.worldbank.org/en/education/From-chalkboards-to-chatbots-Transforming-learning-in-Nigeria

elizabethwithaz.bsky.social•35 days ago

Half the kids got extra services, with a teacher, and the other half didn't, and results are used to show that the technology is effective. In this case, the treatment condition had tons more hours of study time with a teacher, in a subject they otherwise weren't taught. Obviously they learned more.

covingtonedu.bsky.social•35 days ago

But learned more of what? What is this chart showing? Again, because "AI skills", digital skills, & English are all grouped together there's no way of telling how "learning" is defined here.

elizabethwithaz.bsky.social•35 days ago

Yeah, that's what I'm saying. The goal is (supposedly) to learn English skills, but they introduce a new skill to the treatment group and not to control, mix up testing the new skill with English. Of course treatment learned more about the thing the control kids weren't taught, but it means nothing.

martyolliff.bsky.social•35 days ago

But, but, but . . . the TechBros only started advocating for Bloom’s solutions when they had a tool by which they could extract rents.

stanislavfort.bsky.social•35 days ago

There is the practical problem of teachers being expensive. AI is way cheaper. No wonder it looks like a more plausible approach to many.

overit22.bsky.social•35 days ago

Is it really though? The $ spend on development, maintenance, hardware, energy/ water/ environmental issues vs a well educated teacher? Not to mention the difference between simply getting a factual answer (hopefully correct!) to a question vs developing problem solving, process of writing, etc?

bmorebaker42.bsky.social•35 days ago

I don’t need a small city’s worth of water and electricity to answer a kid’s history question

covingtonedu.bsky.social•35 days ago

If a human being required over 7 gallons of water to answer 52 emails, we'd call it a medical emergency.

breweredu.bsky.social•35 days ago

Now that we can monetize 1:1 without enabling education to be a career where teachers are valued, it's clear that 1:1 is the answer!

Teachers: kids need personal feedback, we need to hire more teachers.

Techbro: what I'm hearing is that I need to make a machine to replace you that we can control.

lynnelainece.bsky.social•35 days ago

It has never been about what research says or best practice. It’s always been about control. And don’t forget that teaching is a female-dominated profession.

Comments

Posting Rules

Reply