Uganda will sooner compete in software than agriculture

Soil Quality map source: Food & Agriculture Organisation

The economic principle of comparative advantage is the reason you pay for more things than you make or do yourself, instead of everyone farming their own food, making their own clothes, teaching their kids algebra etc we specialise and pay others to make or do things for us. The same principle behind this specialisation applies to nations. Everyone is better of specialising in what they do more efficiently.
If you're Ugandan this might make you think we should double down on agriculture because it is what we do best. While agriculture might be what Ugandans do best compared to other things Ugandans do, globally we are average at best and at worst, we kinda suck at farming.
The image above shows that our soil is average at best, and some of the most lucrative agricultural export markets already have the best soil. In terms of  productivity we aren't really stellar at anything. In coffee our biggest export, we are the 8th or 7th in coffee exports by volume in the world, by revenue we were around 20th in 2015. We are among the top 3 producers of bananas  but when it comes to exports, we were about 73rd in 2015.

Of course at the moment we are probably even less competitive when it comes to computers and software. However, imagine you had 1000$ to spend on educating a Ugandan high school graduate ad your only choices were agriculture or software engineering. For 1000$ if you go the software route, you could buy the student a laptop, a data plan, a smartphone and sign him up for a bunch of online courses.
At globally competitive levels though, software engineering is ultimately less capital intensive than agriculture. It's much easier to get a 10X return on a one year investment in a computer and internet connection than it is to get the same for the same investment in agriculture. From Kampala, a developer can communicate and find a client in Manhattan. With cloud services, a developer in Uganda has access to almost as much IT infrastructure as one from a developed country. The Ugandan developer though, can easily be much more cost efficient.
The typical entry level annual salary for a developer in the US is about USD 50,000. A similar entry level developer in Uganda probably earns a 10th of that or less. While the environment might give the american based developer an edge, it doesn't make him 10 times better. In any case, the amount of training it would take to get an entry level Ugandan developer to the level of entry level developers anywhere else can be measured in days or at most weeks rather than months or years. Bottomline, it costs far less to create a world class developer than it does to create a world class farmer.While agriculture can and has benefitted Uganda a lot, it might not be what leads us deep into middle income status and it might have to take a smaller role with time.