Sabitlenmiş Tweet
This is sad !! . Uganda 🇺🇬 let’s do something
Hope Moussi@Hope_Moussi
The white space across Africa is glaring. It’s time for us to lock in!
English
Rugaba🇺🇬
1K posts
Rugaba🇺🇬
@rugaba98
|| Founder @sangyinmedia || AI Consultant ll Fmr. President - Uganda Students Association In China (USAC) || 3x Pride of Africa - Asia Awards Winner .
Uganda & China Katılım Temmuz 2018
1.1K Takip Edilen866 Takipçiler
Why not an AI weather prediction system. It’s much more efficient.
UBC UGANDA@ubctvuganda
Kampala launched a digital app to enhance early warning and response to flash floods. #UBCNews | youtu.be/XwLt4NWdslY
English
Rugaba🇺🇬 retweetledi
Gen. Amin Dada's views on Israel and the Middle East
English
@ReachDrMuganga & @VUKampala , you nailed it 💯💯on this .
Valee 🔫@chills_valee
Marketing is evolving and Victoria University is adjusting 😂🙌
English
I honestly don’t rate much of llama models . However, my preference list is Claude , Chat GPT , Deepseek , Gemini & Grok .
Felix Kitaka@TheMutambuze
Stop using Meta AI. Whenever I ask people what AI they use and they say Meta AI, I’m honestly a bit surprised. Because Meta pushes it so heavily through WhatsApp, it’s the very first encounter with AI for most people. But the truth is, you barely scratch the surface of what artificial intelligence can actually do if you stop there. Meta AI doesn't give you a true sense of the current state-of-the-art tech. If you want a powerful free option, just use your Gmail account to try Gemini. If you're willing to pay a monthly subscription for top-tier performance, upgrade to ChatGPT or Claude. Don't settle for the basics. Be blessed.
English
Rather have 8 Official languages. English, Kiswahili, Luganda, Luo , Lusoga , Runyakitara , Runyankore & Ateso . For National languages ( all the indigenous languages should be included) . That’s far enough.
Patro Uganda@PatroUganda
Should Luganda be made the national language of Uganda?
English
English
Meet Vian Kamanzi a student of International Business at @OfficialMubs
Vian is building an AI-powered tool designed to help thousands of university students collaborate more effectively. And he's doing it right now as a student.
This is exactly what AI Studio Uganda was built for.
Not AI for AI's sake. But putting the power of building into the hands of students, problem-solvers, and everyday people who don't yet know they're builders.
The next generation of African innovators isn't waiting for permission. They're already building. 🇺🇬
youtu.be/xutsY0mvZf0
@nssfug @ACCA_Uganda @Absa @ganafaR @1lamerckkavuma
YouTube
English
Rugaba🇺🇬 retweetledi
At the AI Innovation Demo Day 2026, 9 startups pitched AI solutions tackling challenges in healthcare, security & public services.
🏆 Grants were awarded to GRITAP AI, LiverAI, Sukali Check, Law Lens, Simu & Sentinel.
💰Feyti Ltd, Impala Health & Phronesis each received UGX 50M through the AI Health Incubation Grant.
Learn more: shorturl.at/CZgMa
#AIInnovation | #AIinAfrica
English
Rugaba🇺🇬 retweetledi
We’re mobilizing resources for the Uganda Media Centre.
Big thanks to PS @azawedde and the @MoICT_Ug crew; ST @rggoobi, Budget Director Hannington Ashaba, Brenda Nantumbwe, and the @mofpedU team—they totally get why funding can’t wait.
..plus @NITAUgandaED for jumping in with those essential computers.
We’re all about teaming up with the private sector.
We’ll champion creatives, spotlight opportunities for the youth, and stand up for Uganda without holding back.
I will prioritise meeting potential partners.
Time to band together so Uganda gets the respect it deserves.
English
Rugaba🇺🇬 retweetledi
We need to be serious about asking for details when people make claims like this and it starts to build virality.
Sure, there is a chance they are doing what they claim, but there's almost no technical detail here about the capability of this model, how it was trained, and who did the training.
> The site calls a Supabase backend and returns JSON all at once, not streamed. It does not transparently call its own model.
> It returns reformatted model outputs with what looks like backend orchestration to search/tools and an LLM.
> They're sending conversation history that makes it look more like app-layer than a model-serving API.
At a minimum, they aren't exposing a model inference endpoint. They're sending messages to a backend that then feeds it into search/LLM elsewhere then spits a result back they modify and serve. They still could be forwarding this chat-wrapper to their own LLM but there's no evidence.
It's much more likely this is either:
1. A wrapper/orchestrator on top of an existing LLM API with custom prompting.
2. A wrapper/orchestrator on a fine-tuned open model hosted elsewhere.
The onus is on this team to clarify their claims and happy to update with technical details.
Veta Origin@vetaorigin
Global AI wasn’t built for Africa. So we built one that is. Powered by an African LLM.
English
Rugaba🇺🇬 retweetledi
Computer science is gradually returning to the domain of physicists, mathematicians, and electrical engineers as large language models automate much of what we currently call software engineering.
The field’s center of gravity is shifting away from manual code writing and toward deeper theoretical thinking, mathematical insight, and systems-level reasoning.
English
English
Am building safe and policy aware agentic sdks for Africa, anyone interested in collab or any forms of partnership inbox, I already have a team and company, so I expect serious people only, also this am not talking of employing anyone am looking for fellow technical fellows we can solve the puzzles and research together with. Africa needs it's own power and moving forward power is data and contextual intelligence. @NKaliisa if you can also angel invest in a preseed then we can change the narrative together with practice not words.
English
We can speak about AI and hold conferences about it, but if we as Africa don’t take the practical lane, nothing will ever change. We will remain consumers not producers.
English
@CarolPrim3 @gad_rogers @MinofHealthUG @DICTSMakerere @Makerere @STIsecretariat The QR code doesn’t work @CarolPrim3
English
Excited to announce the AI Innovation Academy Demo Day 2026. We're spotlighting talented pre-seed AI pioneers crafting practical solutions for everyday problems.
This event unites entrepreneurs, guides, collaborators, funders, academics, and business experts to witness bold ideas transitioning into impactful startup ventures.
#AIInnovation | #AIEntrepreneurship
English
@weskambale @TimKalyegira I beg to differ @weskambale . Maybe in Africas case . But in Asia definitely not . We will discuss more when we meet.
English
recently, when i was tellin' @timkalyegira that ai still has a long way to go, i didn't know anthropic would drop such a fascinatin' study
from the radar chart, two things are obvious to me:
theoretically, ai coverage across occupational categories is remarkable, a i said it has improved since 2022/23
but the observed coverage, what's actually bein' deployed and used, is only a fraction of that
ai as it exists today is still relatively new, usin' it effectively requires skills like prompt engineerin', agent buildin', and system integration that most organisations don't yet have
there's also europe's regulations that are slowin' enterprise adoption, and africa's infrastructure limitations that make meaningful adoption unlikely anytime soon
now look at what's barely covered under both theoretical and observed:
healthcare support, production, transportation, agriculture, construction
all these are the categories that run the physical world
the most exposed occupations in the study are programmers at 74.5% (not system-level), customer service at 70.1%, and data entry at 67.1%
notice anythin'?
they're all text- and data-processin' jobs, and that's exactly what llms were built for
generatin' a 20-page report in 30 seconds, which Tim finds fascinatin', is not integration into the physical economy
it's not a workin' telemedicine system. it'is not a self-drivin' logistics network. it' is not an autonomous production line
those require real-time decision-makin', hardware integration, and regulatory clearance that current llms simply don't have
and the people buildin' these systems understand this, which is why they do not celebrate llms generatin' reports as Tim would want us to do
when ai hits 60%+ observed coverage in those untouched physical-economy categories, then we can have a different conversation
for now, i still maintain, we're not close.. ai still has a long way to go
the study: anthropic.com/research/labor…
English
@indabaxug @indabaxug how does one register for your conference attendance
English
@solomonking @SueNsibirwa @elementaledge @solomonking would you please mind checking your inbox 📥. Thank you
English
This is 100% true.
I'll always be grateful to @SueNsibirwa for being our first major client for NVGOR8 (later @elementaledge) way back in the day.
That trust from a major brand (MTN) opened up so many doors for us to become a big production studio.
Nikhil Krishnan@nikillinit
I'm starting to realize that 4-5 key people will change the trajectory of your entire career or business. Some archetypes this can look like: - A fantastic boss/manager - A champion at a first customer that opens doors for you - A colleague that vouches for you to get a job that's out of your league - A mentor that gives you honest feedback of your current business and what you need to change Finding those 4-5 people is way more important than being popular, having followers, etc. Find whatever spaces that will have the highest likelihood of those 4-5 people being there.
English
Rugaba🇺🇬 retweetledi
If you’re a user of AI, delete GPT.
Anthropic turned down a deal to allow their models to be used by the government for surveillance and autonomous weapons.
Open AI took the deal.
Sam Altman@sama
Tonight, we reached an agreement with the Department of War to deploy our models in their classified network. In all of our interactions, the DoW displayed a deep respect for safety and a desire to partner to achieve the best possible outcome. AI safety and wide distribution of benefits are the core of our mission. Two of our most important safety principles are prohibitions on domestic mass surveillance and human responsibility for the use of force, including for autonomous weapon systems. The DoW agrees with these principles, reflects them in law and policy, and we put them into our agreement. We also will build technical safeguards to ensure our models behave as they should, which the DoW also wanted. We will deploy FDEs to help with our models and to ensure their safety, we will deploy on cloud networks only. We are asking the DoW to offer these same terms to all AI companies, which in our opinion we think everyone should be willing to accept. We have expressed our strong desire to see things de-escalate away from legal and governmental actions and towards reasonable agreements. We remain committed to serve all of humanity as best we can. The world is a complicated, messy, and sometimes dangerous place.
English