💡 First Day of School-GPT…

Hi there,

Too-woke AI? The latest in the “should-AIs-be-woke?” debate is Goody-2, an AI so ethical it won’t answer any questions at all. Go on, just try to ask it something inoffensive.

In this Open Letter:

• Smarter EdTech: The business case for AI in schools.

• Calls from your X, lost traffic fines and SA’s first moon base.

• Killer marketing: How to build a startup content strategy.

• Where the doctor belongs: The results are in.

• Free business stuff: Share this and get cool tools.

TRENDING NOW

The Case for More School-GPT

UNESCO recommends there be one teacher for every 20 kids in a classroom.

And, while the average ratio of SA public schools is around 30:1 (crowded but manageable), 15% of primary schools have over 50 pupils per class.

In fact, half of schools have class sizes exceeding 40 students. And recent reports are that the average class size for grade 6 learners is 61 in Limpopo, 59 in Mpumalanga and 54 in KwaZulu-Natal. 

Bless those poor teacher’s souls!

All about maximising efficiency

And there’s a reason UNESCO proposes that ratio:

• Increased Individualized Attention

• Improved Learning Outcomes

• Enhanced Classroom Management

• Greater Teacher Effectiveness:

However, getting to this ratio could be extremely expensive. 

Getting our ratio in public schools down to 20:1, means we need another 200’000 teachers to serve the ±12m learners. 

And, at an average income per government teacher exceeding R500k (if you include all the benefits) per annum, just the salaries would increase the government’s education spending of R368 billion by another R100 billion.

SA simply doesn’t have the money for this.

AI assistants could help

Some schools are making use of classroom assistants or tutors that can help reduce the load on teachers at a reduced cost – they earn less per hour and can come in for shorter bursts where needed. 

With the advancements in large language models (LLMs like GPT-4 etc), however, you could use AI to supercharge a teacher and their capability to deliver high-quality education even with larger classes. 

This could have a great impact on understaffed public schools. 

But it would likely take years for the government to roll it out.  

And that’s why a good place to start might be the private sector…

The Business Case for AI in Private Schools

Private school group ADvTECH reports that 84% of their building capacity is currently utilised – efficient, for sure, but if teachers can handle bigger classes with the support of AI, there is money to be made. 

Even if AI tutors can increase classroom capacity by just 1%, it could increase annual revenue by R40m for the group (assuming an average annual school fee of R100’000) – significant. 

They currently have just under 40’000 students enrolled and it's been growing at 10% per year – highlighting a healthy demand. 

And that’s exactly why they have been looking at using AI.

ADvTECH schools use ADvLEARN a customised platform to provide personalised learning paths and use adaptive technology to give students a bespoke learning experience while improving their understanding in core areas. By helping teachers pinpoint where individual pupils are struggling and using data from locations around the continent to improve the curriculum and upskill teachers, it’s a winning formula.

But they are not the only ones…

A startup doing something interesting in this space is Mindjoy. It’s like School-GPT: They aim to help schools develop their tech and STEM talent amongst teachers by providing STEM teachers with their very own customisable AI Tutor to personalise and enhance their students’ learning experience. 

You can give it a personality and avatar and get kids to not only learn from it but give it insights (which go to the teacher) on how well students understand topics. Call your bot Einstein and get it to explain the theory of relativity to a 10-year-old – nice.

Have a look…

As AI continues to advance at breakneck speed, more and more industries (just like EdTech) will be able to leverage their deep-rooted domain knowledge and create personalised AI-powered solutions to add value to users and their industries. 

Not only will this have a positive impact on education, but there sure is money to be made. We’re watching this space.

IN SHORT

🇿🇦 Scientific Champions. Two aspiring young scientists from SA won big at the 2024 Taiwan International Science Fair walking away with a Third Award in the computer science and information engineering category, a Fourth Award and the Viewer’s Choice Award in the behavioural and social sciences category.

📞 Calls from your X. Soon Elon Musk will ditch his phone number and only use X for texts and audio/video calls in a bid to make X the one app to rule them all. Just hope he has his 2-Factor Auth set to Email.

♊️ Gemin(A)i. Google’s AI tool Bard has officially become Gemini with the launch of Gemini Advanced with Ultra 1.0 which is already available in 150 countries & territories including South Africa.

👮‍♀️It’s Not Fine. Looks like you’ll never get a traffic fine again. The SA Government wants to force traffic departments to use the Post Office to issue fines as part of the proposed Administrative Adjudication of Road Traffic Offences (Aarto) Act said to be implemented on 1 July this year.

🌝 To the moon. SA will work with China, Russia and a bunch of other countries to build a lunar base on the moon as part of The International Lunar Research Station (ILRS) project.

🤑 Raising funding. Sam Altman is on the hunt for $7 trillion in investment. The OpenAI head honcho is apparently in talks with officials in the UAE to present his plans to build a whole lotta new chip factories (OpenAI guzzles up a LOT of computing power).

👨‍💼 User Words. Are you a techie battling to explain your value proposition/what your product is about? Business comms author Chris Fenning is doing a free talk on turning “Geek Speak to Street Speak” at Innovation City this Wednesday (14 Feb) at 1pm. Should be great for learning to chat with users and stakeholders and feel less awkward about it.

BUILDER’S CORNER

How to Create a Killer Startup Content Strategy

You’ve got a product (or on its way!) and it’s time to start creating some buzz around it, building an audience and activating potential users. Great!

But now comes the thing: What content exactly are you supposed to do?

I mean, it’s a busy space; web, socials, YouTube, podcasts – there’s so much going on already. How do you stand out?

Nothing drains your marketing budget (and your audience’s excitement) faster than dull, boring copy-pastes of the same old thing everyone else is doing. So it’s time to think way outside the box…

Unlocking the “Unique” in a Crowded Space

1. Forget your brand & pretend you’re Rupert Murdoch

The biggest mistake brands make around content is they start with questions like: What do we (the brand) want to tell people? Content is just like a product – it’s all about the user and their journeys, not yours.

Rather: pretend you’re an editor who was just hired by a big media company to create a new publication/podcast/TV show etc. targeting your audience. What commercial content product would you build to attract and engage them? Make that your content strategy.

2. Give them something they can’t find anywhere else

Go find out 1) what your audience really wants to know/see and 2) what information is not available out there on the net or socials yet. And build that into a strategy.

Example: I once had to do a strategy for a boat dealership. All the competitors were posting about engines and boats. We went a step further and said: What do people use boats for? Fishing, of course. So we became the first boat brand to build content around fishing and fishing competitions instead, and were rewarded with tens of thousands of active, engaged users and zero competition for our content (because they couldn’t get it anywhere else).

3. Or just ship it on a brand-new channel

Ok, so some topics are done to death. In most places, but usually not on all channels. So you can use the existing channels as your resource and ship the content on a new channel (that you know your users would really prefer.)

Example: Yeah I lied about being the first to talk about fishing competitions for that boat brand. You could get lots of fishing news at that time; the problem was it was only in printed magazines. So all we really did was take that info and publish it online and on the socials – where people actually wanted it – and boom! Content success.

You can do it too. Just look at what’s available on blogs but not on YouTube yet. Or deep-dives on podcasts that are not on carousels or video yet and give it to your people.

BONUS: Thinking of your content as a separate “product” is very useful, because it forces you to ask: How’s this gonna make money (or pay for itself)? It reminds you to include your organic traffic in your marketing channel analysis and physically check: How many new users is it bringing in, at what cost (CAC), and what’s the conversion rate?

Remember: The ultimate goal of content is for organic to eventually replace or overshadow your paid traffic, so you can lower your CAC and make a profit.

Got startup content and marketing insights? Hit reply and let us know (and maybe you get featured here, too).

THE RESULTS

We asked where and how you see your doctor, and good old face-to-face is still the game…

🟩🟩🟩🟩🟩🟩 🧑🏻‍⚕️ In-person at their offices (76%)

⬜️⬜️⬜️⬜️⬜️⬜️ 🏥 At the clinic (5%)

⬜️⬜️⬜️⬜️⬜️⬜️ 📱 On my phone (3%)

⬜️⬜️⬜️⬜️⬜️⬜️ 💻 On my laptop (3%)

⬜️⬜️⬜️⬜️⬜️⬜️ 🚑 Only in emergencies (hospitalisation) (3%)

⬜️⬜️⬜️⬜️⬜️⬜️ 💪 Haven’t seen a doctor since 1999 (10%)

Your 2 cents…

FOR THE MEMES

Got startup memes? Send them our way or tag us on socials.