Hi! Welcome to the 19th edition of the TomorrowToday newsletter.

We’re here to decode the AI chaos so you don't have to. Think of us as your friendly neighbourhood tech translators - we cut through the chaos, translate the jargon, and spotlight new AI tools that matter for founders, builders, and curious minds.

Buckle up, because the future's moving fast and we're here to make sure you don't get left behind! ⚡

If you enjoyed today’s newsletter, please forward it to a friend & subscribe by following this link.

_{~7 mins read}

/Pricing /Accessibility /Africa

OpenAI just released ChatGPT Go, and it's a big deal for South Africa. Sam Altman and the team brought it to Africa with rand-based pricing at R149/month.

Let's be honest: this changes things. For the first time, South Africans get access to GPT-5 without USD conversion fees eating into the cost. R149 stays R149. That's significant.

What ChatGPT Go gives you that the free version doesn't:

The free version of ChatGPT is limited. You get basic access to older models, limited message counts, no file analysis, and no memory between conversations. You're working within tight constraints.

ChatGPT Go flips that. You get GPT-5—OpenAI's best reasoning model. That means better analysis, smarter answers, and actually useful results when you're trying to solve real problems. You get 10x higher message limits so you're not constantly hitting walls. You can upload files and have ChatGPT actually analyse them. You get advanced data analysis built in. And the model remembers context from your previous conversations, so you don't have to re-explain everything every time.

For someone using ChatGPT regularly, the free version becomes frustrating fast. You hit limits constantly. The model struggles with complex reasoning. File uploads either don't work or require workarounds. ChatGPT Go removes all that friction.

Why R149/month actually matters:

This is the price of three beers. For students grinding through assignments, freelancers bidding on projects, or small business owners trying to scale without hiring—this is access to enterprise-grade AI at a price that makes sense.

The free version is a taste. ChatGPT Go is actually useful, and now a large group of people who cant afford the $20 subscription to OpenAI can get access to premium features.

/Pricing /Coding /Performance

Anthropic just gifted developers one of the best and cost effective models when they released Claude Haiku 4.5 this week. Five months ago, Claude Sonnet 4 was frontier. Today? Claude Haiku 4.5 delivers the same coding performance at one-third the cost and twice the speed.

Let that sink in.

Haiku 4.5 costs $1 input / $5 output per million tokens. Sonnet 4.5 runs ~$3 input / ~$15 output. GPT-4o sits around the same price as Sonnet. Haiku wins on both price and speed.

But here's what matters: Haiku 4.5 doesn't just match Sonnet 4—it beats it at certain tasks. Computer use, for example. It's not "almost as good." It's legitimately better at specific things.

This matters for real work. Chat assistants that need instant responses. Customer service agents handling hundreds of conversations. Developers pair programming. Haiku 4.5 handles all of it faster and cheaper. The benchmarks prove it's close to frontier performance without the frontier price.

/Enterprise /Automation /Workflow

Most people using Claude won't implement this. But if you do? Everything changes.

Claude can now use Skills to improve how it performs specific tasks. Think of Skills as teaching Claude exactly how you work, not how some generic AI thinks you should work.

Normally, Claude generates output. You review it. You reformat it. You adjust it to match your brand. You send it. Rinse and repeat. With Skills, you upload your brand guidelines, company standards, exact format preferences. Claude does it right every single time. Automatically.

Skills are portable (work everywhere—Claude app, Claude Code, API), composable (they stack together), efficient (only load what's needed), and powerful (they can run actual code).

You can build brand-aligned slideshows by uploading your brand guide and having Claude create presentations automatically. Excel workbooks with formulas and your company structure. Custom documents formatted exactly how you need them. Sales playbooks, customer service responses, compliance checklists.

Claude automatically identifies which skills are needed and uses them without you asking. You don't upload and manually select every time. It just works. For teams and enterprises, this is transformative. Instead of three hours reviewing and reformatting AI output, you get work-ready results in seconds. Your brand never drifts. Your processes never vary.

In this section, we aim to spotlight an incredible AI tool or use case and guide you on how you can try it.

This week's challenge: Turn your lecture notes into an interactive study guide with NotebookLM 📚

You know that feeling? 50 pages of lecture notes. A textbook chapter. Zero motivation to re-read it all.

What if your documents could just explain themselves? That's NotebookLM.

Upload a PDF, Google Doc, or any document. NotebookLM reads it and generates audio overviews, interactive Q&A, slide summaries, study guides, podcast transcripts, video summaries. Basically everything except writing the exam for you.

You can study for exams by pasting your notes and uploading the textbook chapter to get a study guide tailored to both. Upload your product manual to create a 24/7 chatbot that answers customer questions. Take your company handbook and generate interactive training content. Upload 5 academic articles to get synthesised insights and comparisons. Turn confusing contracts into plain-language explanations.

How to do it (5 minutes):

Notebook LM is not just a powerful tool for students. Anyone who needs to learn anything or who wished that they can talk to documents will benefit from the frontier tech that Google is offering for free with notebooklm. It is genuinely so good that they can easily charge R400pm for access.

You spoke, we listened. “AI in Enterprise” is here to stay. In this section, we're spotlighting real businesses using AI to solve actual problems.

This week: reMarkable scaled customer support without hiring more people 📞

reMarkable makes premium paper tablets. They've been growing fast—nearly $500 million in revenue, expanding into B2B. Everything is exciting. Then the problem hits.

Customer inquiries explode. Black Friday comes. Cyber Monday arrives. Their support team is drowning. They need to scale support, but hiring 20 new people just for peak seasons doesn't make financial sense. Sound familiar? Every growing company faces this.

reMarkable deployed an AI customer service agent called "Mark" using Salesforce Agentforce. They built it and launched it in three weeks. Mark isn't a chatbot. It's designed to understand real customer conversations, handle context, and know when something's too complex for AI. It works 24/7. It learns as it goes.

Mark has handled 25,000+ customer conversations, resolved 35% of inbound inquiries autonomously without human intervention, with rising NPS scores and improving deflection rates every week. They scaled through Black Friday and Cyber Monday without hiring extra staff. That's not just AI deflection. That's better customer experience with fewer humans.

Three weeks from decision to live. 35% fewer support tickets in your queue means Mark handles routine stuff that doesn't need humans. Customers are happier, not frustrated. Mark is sometimes better at answering simple questions than tired support reps at 2 AM. When Black Friday hits and inquiries spike 300%, Mark handles it. No recruitment, no training, no costs spiralling.

reMarkable didn't fire their support team. They augmented it. Humans now handle complex, high-value conversations. Mark handles "Where's my order?" and "How do I reset my tablet?" and "What's your return policy?"

Humans are happier (meaningful work). Customers are happier (faster answers). The company scales profitably. This is the blueprint for any company growing fast while support costs climb.

AI is full of jargon, and we’re here to decode it. Each week, we’ll give you a plain-English definition of a buzzy term you’ve probably seen (but never fully understood).

Thanks for reading TomorrowToday! We’d love to hear from you:

➡️ What would you like us to cover next?
➡️ Have a tool or topic we should feature?

We’re building this with (and for) you. 🚀
See you next Tuesday 👋