One of my goals this year was to organize my long form writing in a better way and to that end, I'm starting a substack publication I'm calling The Long Tail.
I'll be primarily covering African venture investing and tech trends in the AI era with broader thoughts on frontier markets as applicable.
I hope it drives conversation about our ecosystem, highlights under-appreciated teams, and challenges me to update my priors.
Link to subscribe in the comments.
I finally put all my articles and infographics in order on https://t.co/EroURcijje
I am also starting a short email note alongside it. The first one goes out TOMORROW and connects three recent pieces:
1. How tokenized deposits actually settle
2. Corporate treasury and precautionary cash
3. Why instant settlement runs into prefunding
Writing them back to back left a gap around bank funding incentives that I did not cover in any of the three, so I wrote a quick note on it for the list
It asks for an email to get in, which also puts you on the list for that first note
Attacks will accelerate from here on - We are there: Six months ago I said Mythos like capability will be widely available. It's happening. A critical vulnerability, a social engineering attack that looks innocuous, or daisy chaining publicly available vulnerabilities and conducting a sniper strike is all being witnessed in the wild. Incident responders have never been busier.
Some key points:
- attacks are not becoming more sophisticated, it's the same attacks, faster and by a lot more people.
- attack path discovery which took trial and error and multiple attempts is now trivial, open source LLMs are here to help
- you are as secure as your supply chain, insecure down or upstream players who have access to your infrastructure could be the next attack path, marketplaces and supply chain is an easier vector
The increase in activity is a result of more people becoming better at hacking due to the ease of being able to attack. Had not contemplated that in my calculus.
Please:
1. Test your infrastructure for accessible vulnerabilities and misconfigurations
2. Fix your code and demand your vendors do.
3. Old infra, or insecure infra at the perimeter needs to be updated first
4. New AI tools, especially coding need a security framework. Your developers will do more with code, new skills, and more automation. Don't forget the coding models have trained on human code.
Stay safe this Halloween :)
Cybersecurity will be the ultimate “app” for AI and easily consume more than half of all tokens in the coming year
It will single-handedly justify all of the current compute buildout and we will not have nearly enough
There's a clear pattern to some of the more exciting African AI startups we've been backing. It's not the only pattern but it is clear:
> They attack a *hyper-narrow* wedge into giant industries. Energy, defense, critical minerals, construction, etc. GDP moving categories.
> They find PMF through labor replacement initially. The continent is ripe with insanely inefficient process. We are now in a world where AI startups can offer 10-100x cost savings. This shifts the bottleneck from sales to token efficiency.
> They are then capable model manipulators. They fine-tune open-source, they orchestrate, they run evals. This cannot be faked. I love how clearly we can judge technical depth these days. With it, they can dramatically reduce the token bottleneck.
> They have a vision of owning the context moat for those industries. Not just generic 'data' but the operational context for what drives the real economy, but at machine speed.
> They are small teams but fully agent- augmented, meaning they need less money to do bigger things. I cannot stress how different AI obsessed teams are vs. those even 3 to 6 months behind.
There's some aspects about this pattern that are frontier market exclusive, including just how quickly we're seeing young, non-insider founding teams partner with legacy giants in otherwise mature sectors.
This means they are getting proprietary product/model/data reps in at a higher pace than their peers in other regions. The result is faster global viability in niche but critical industries. This era will be Zipline x1000.
Several of these companies will announce rounds in the coming weeks/months. Very excited for this generation of African founders.
@AppsolutelyxD The incumbents have zero in-house AI manipulation capacity.
Running unmodified open models within your systems is both dangerous and inefficient
The context moat goes beyond a single customer's data, to more of an industry wide context layer while respecting data propriety.
Our last investment was in set of exceptional African founders building AI-native cybersecurity for sovereign deployment on the continent and beyond.
We had a huge debate on market size and ultimately decided that an incoming shift in the nature, surface area, and frequency of cyberattacks meant that governments and multi-national budgets on the continent will have to expand considerably to harden critical assets.
TBD but seems directionally correct given current events.
i recreated THE ENTIRE WORLD in minecraft at 1:1 scale. every building on earth, including your house!
to do this, i trained a SOTA world-to-minecraft generator on maps data, complete with furnished home interiors!
play and check out your neighborhood: https://t.co/m4vk9WqBZp
Today we announced Lemon Squeezy is winding down.
Building Lemon Squeezy and supporting the businesses created has been the privilege of a lifetime. I’m incredibly grateful to everyone who trusted us along the way.
Everything we learned has helped shape Managed Payments at Stripe, where we’re excited to continue the work of making online selling easy. I know this transition will be disruptive and not all Lemon Squeezy users can use Managed Payments. I don’t take this lightly.
Thank you to everyone who built with us and helped create Lemon Squeezy.
It doesn't matter where the vulnerability sits. If it's harming your customers, it becomes your problem. Proactive AI is the only viable strategy.
Cyber capability in Africa coming into focus faster than even our own thesis.
Now the critical infra vulnerabilities...
In early October, a wave of cyberattacks, leading to fraud losses, hit several Nigerian banks and financial institutions, potentially costing billions of naira in customer funds, according to several sources with direct knowledge of the matter.
The attacks, the sources noted, affected institutions that rely on external technology providers to run critical financial services, including core banking infrastructure. In one incident, a vendor providing core banking services to financial institutions was reportedly compromised at the system level on October 1. The vendor is one of Africa’s biggest fintech companies.
I’m looking to speak with someone who has experience building communities for B2B companies.
I have an idea I’d love to get some expert input on and would be happy to pay for a few hours of their time.
If this sounds like you, or you know someone who might be a good fit, please DM me.
African AI ambitions needs to pay attention to @SakanaAILabs.
They've shown leadership in 'resource constrained' deployment that'll uncover many lessons for the fast followers.
A diff level of 'resource constrained' but there are primitives worth paying attention to.
It takes one person to believe in you.
For me, one of those people in my life was @HansMorrisSF 👇
2017 - moved back to Tanzania on a one-way ticket to build a company at 24. pivots begin.
2018 - spinning my wheels. building things that didn't pick up.
2019 - we finally got into @ycombinator on my sixth attempt. One week before Demo Day, my co-founder quit. I had to tell all our investors before they wired that i'd be building this company without a co-founder. (back then, this massively concerning)
Hans was the first institutional VC (David Sica had connected us via an intro from @dfsventures - @mrstephendeng@JakeKendall5 who gave us our first check ever)
We met at a small coffee shop in San Francisco. He told me he believed I'd figure something out, and he wanted to take the bet. Hans then invested $500K into nala.
2020. COVID hit, we received cease and decease letter from the largest mobile operator in Tanzania, the whole business pivoted (yet again). Hans was patient and supportive.
2021. We pivoted again. Hans would jump on the phone with me as i was telling him about wanting to get into cross-border payments.
he worked through scenarios, our financial design and the model with us, even when things were really really hard.
2022. Our cross-border payments business started growing fast, and Hans invested in our $10M round.
2024. He invested again in our $40M Series A. Led by @AcrewCapital and the incredible team there!
He's one of the few investors who has visited us in multiple cities and offices. Today, he came to spend time with me in London, to go over the business, financials, and share, as always, a lot of helpful advice.
Hans stood by us through some not-so-(stable)coin times.
Yesterday, one of the largest global cross-border businesses signed a contract to be our customer for global stablecoin payouts.
Thank you, Hans, and thank you, @nyca partners for believing in NALA and our team.
One believer in our ambition can change a company's whole direction. we are still a long way to go for where we want to be, we have big big plans. I wanted to give a cheers to people like Hans Morris 🙏🏽
There's one truth in xborder trade in many African countries like Nigeria: the dollars will flow.
> This floods the local compliance path with cost and disincentive, clearly aiming to mold stables into the shape of the current fiat FX structure. But crypto's malleability is self-determined.
> This places even more value on offshore USD fiat as an intermediary buffer. But then there's this:
Section 11.10.8: Virtual assets cannot ordinarily be used for offshore settlement, prefunding or netting of FX and remittance obligations without approval.
> This attempts to flood that channel with reg risk, but that's been historically hard to enforce. This will result in higher prices, but the value continues to accrue to the same:
1. Those with reliably replenishable fiat dollars
2. Those with pre-aggregated local currency demand
Ironically, punishing the transport of USD stables will likely further empower offshore USD fiat.
> Those bypassing the above by effectively shadow-banking for the regulated banks will probably need to see how this provision plays out politically:
Section 11.10.17: The CBN can recognize foreign stablecoins for wholesale cross-border settlement without requiring the usual local issuer arrangement, but only under restrictive conditions and not for retail distribution.
> More challenger startups enter as costs rise, but many won't survive long enough to learn that absorbing risk is not the same as systematically reducing it.
JUST IN: Nigeria’s Central Bank introduces stablecoin licensing rules, banning algorithmic stablecoins, requiring separate licenses for naira and foreign-currency stablecoins, and giving existing operators 180 days to comply.