Category: MIL-OSI

Source: The Conversation – Africa – By Justin Willis, Professor of History, Durham University

Raila Amollo Odinga, who has died at the age of 80, was something of a paradox in post-independence Kenyan politics.

A leader who repeatedly ran for president, he never won – in part due to the 2007 election being manipulated in favour of Mwai Kibaki. Despite this, Odinga will be remembered as a figure who profoundly shaped the country’s politics as much as any president.

The son of a famous anti-colonial leader, he was born into influence. Yet he became bitterly critical of Kenya’s enduring political and economic inequalities, speaking out on behalf of the county’s “have nots”, which earned him a place in the hearts of millions.

He was a fiercely nationalist politician who mobilised support across ethnic lines. But he was also the dominant leader of the Luo community – one of the country’s larger ethnic groups mainly based in Western Kenya – whose voters formed the core of his support.

Having self-identified as a revolutionary, Odinga later proved to be committed to institutional reform and democratisation. His greatest legacy is the 2010 constitution, which attempted to devolve power away from the “imperial presidency”, which he campaigned for over many years.

This was not the end of the contradictions. A leader who often spoke about economic development and deprivation, his agenda was typically more focused on political change. Odinga did so in part because he believed that rights and freedoms would anchor nation-building and development.

Perhaps most strikingly, although he scorned the elite power sharing deals that dominated Kenyan politics – he repeatedly made such agreements himself, often invoking the need for national stability.

Odinga embodied Kenya’s political contradictions, so the impact of his life and death will be debated. This article explores this contested legacy and what it means for Kenya’s future.

Early years

Born in western Kenya on 7 January 1945, Odinga – popularly known as Baba (father) – was the son of Jaramogi Oginga Odinga, the redoubtable community mobiliser who was a thorn in the side of the colonial state. Oginga famously insisted that he and other nationalists would make no deals with the British until Jomo Kenyatta was released.

When Kenyatta became prime minister in 1963, and later president in 1964, Oginga became Kenya’s first vice-president and minister of home affairs. However, he fell out with Kenyatta in 1966 over the government’s failure to overturn colonial inequalities. This meant that the Oginga family was excluded from the country’s powerful political elite. Oginga spent the following decades in and out of detention.

Raila Odinga spent his early years in Kenya before leaving in 1962 to study in East Germany. Returning in 1970, he became a university lecturer. Later, he joined the government standards agency – a job he lost abruptly in 1982 when he was linked to a failed coup against Daniel arap Moi. Charged with treason, he was detained until 1988, when he became active in the growing opposition to Moi’s rule. He was detained twice more during the turbulent years of protest that followed and fled briefly to Sweden.

Odinga returned before Kenya’s 1992 elections, the first multi-party polls since the 1960s, siding with his father when the opposition split. Aided by that division and state manipulation, Moi won, but Odinga’s role confirmed his status as a major political figure.

Blazing his own trail

When Oginga died in 1994, Odinga sought to take over his father’s party but, defeated, left to form his own. He ran for president in 1997, which Moi again won against a divided opposition.

When Moi did not seek re-election in 2002, it seemed Odinga’s moment had come. However, after briefly supporting Odinga as his successor Moi ultimately decided to back Uhuru Kenyatta, son of Jomo. In response, Odinga threw his weight behind Mwai Kibaki, a move which was critical to Kibaki’s victory in 2002.

Odinga’s support for Kibaki was conditional on major constitutional and political reforms. Yet where Odinga had expected widespread constitutional reforms to devolve power away from the executive, Kibaki offered limited changes. Refusing to simply prop up the administration, Odinga successfully campaigned against the government’s flawed draft constitution in the 2005 referendum.

Once again, Odinga seemed on the brink of power: he led a broad coalition into the 2007 elections on a promise of fundamental change. Early results put him ahead of Kibaki in the elections – but then Kibaki was declared the winner in a hasty process that raised widespread suspicions of malpractice and triggered Kenya’s greatest crisis, including ethnic clashes and state repression.

A power-sharing deal brought the violence to an end and made Odinga prime minister in a government of national unity. He focused his energy on political reform and constitutional changes, as well as other long standing concerns. In August 2010 a referendum approved a new constitution that devolved power to Kenya’s 47 counties. The constitution also reformed key institutions including the judiciary and electoral commission and expanded citizens’ rights.

A contested final act

The 2010 constitution remains Odinga’s signal achievement. Certainly, it created the potential for the country to forge a new and more democratic future.

Yet in its aftermath he struggled to find an equally compelling narrative. Constitutional reform had been a long-standing demand that allowed him to mobilise opposition around the promise of a new Kenya. Without this single over-arching “cause”, Odinga’s ability to sustain mass mobilisation became more fragile.

Furthermore, the progressive constitution did not prevent the continuation of older political logics. It proved no barrier against the rise to the presidency of Uhuru Kenyatta and his then deputy, William Ruto, who had faced charges of crimes against humanity at the International Criminal Court.

Odinga faced increasingly difficult choices, particularly after repeated presidential defeats in 2013, 2017 and 2022 amid allegations of electoral manipulation.

These losses convinced some that he would never win the presidency – and not only because of the use of state power to deny him. That recognition, coupled with advancing age and ill health, led Odinga to make compromises once unthinkable, revealing an increasingly pragmatic reasoning in his later years. This was starkly illustrated after the 2017 elections, when – having claimed he was rigged out and led mass protests – Odinga struck the “handshake” deal with Kenyatta in March 2018. This was framed as nation-building but viewed by some as a betrayal.

The handshake led Odinga to stand as Kenyatta’s preferred candidate in the 2022 elections. This backing proved doubly damaging, however. On the one hand, it undermined Odinga’s opposition credentials and lowered turnout in his Nyanza strongholds. On the other, it meant that his loss could not be blamed on a “deep state” conspiring against him.

The difficulties that followed were magnified when, after suggesting the 2022 results had been manipulated by those around Ruto, Odinga agreed to prop up Ruto’s struggling government in March 2025. The formation of what was billed as a “broad-based” administration was presented as nation-building, but critics saw it differently. Coming after mass youth-led protests – first against tax increases and later against corruption, state repression, and Ruto’s leadership – Odinga appeared to some to side with power against the people he once represented.

Not flawless, but consequential

These turns complicate how history, and Kenyans, will remember him – not as a flawless icon, but as a deeply consequential and sometimes contradictory figure. Yet those with longer memories will also understand what led Odinga there.

Imprisoned and tortured under Moi, sold out by Kibaki, and denied victory in 2007, Odinga endured more than a lifetime’s share of misfortune and betrayal. He made his own choices, but rarely under conditions of his own making, and arguably did more than any other Kenyan to make the country’s political system more responsive to its people.

His absence will generate a political vacuum that other leaders will struggle to fill. Ruto was banking on Odinga’s support to win the 2027 elections. He will now have to work harder to put together a winning coalition. Meanwhile those leaders who coalesced around Odinga – including those who depended on him for their positions – will need to decide how they can most effectively mobilise in his absence.

As they do so, Kenya’s leaders will all be operating in his shadow, and in a context in which the country’s marginalised people and communities will feel even less represented by those in power.

– Raila Odinga: the man who changed Kenya without ever ruling it
– https://theconversation.com/raila-odinga-the-man-who-changed-kenya-without-ever-ruling-it-267643

Source: The Conversation – Africa – By Vukosi Marivate, Chair of Data Science, Professor of Computer Science, Director AfriDSAI, University of Pretoria

Artificial intelligence (AI) tools like ChatGPT, DeepSeek, Siri or Google Assistant are developed by the global north and trained in English, Chinese or European languages. In comparison, African languages are largely missing from the internet.

A team of African computer scientists, linguists, language specialists and others have been working on precisely this problem for two years already. The African Next Voices project, primarily funded by the Gates Foundation (with other funding from Meta) and involving a network of African universities and organisations, recently released what’s thought to be the largest dataset of African languages for AI so far. We asked them about their project, with sites in Kenya, Nigeria and South Africa.

---

Why is language so important to AI?

Language is how we interact, ask for help, and hold meaning in community. We use it to organise complex thoughts and share ideas. It’s the medium we use to tell an AI what we want – and to judge whether it understood us.

We are seeing an upsurge of applications that rely on AI, from education to health to agriculture. These models are trained from large volumes of (mostly) linguistic (language) data. These are called large language models or LLMs but are found in only a few of the world’s languages.

---

Read more: AI in Africa: 5 issues that must be tackled for digital equality

---

Languages also carry culture, values and local wisdom. If AI doesn’t speak our languages, it can’t reliably understand our intent, and we can’t trust or verify its answers. In short: without language, AI can’t communicate with us – and we can’t communicate with it. Building AI in our languages is therefore the only way for AI to work for people.

If we limit whose language gets modelled, we risk missing out on the majority of human cultures, history and knowledge.

Why are African languages missing and what are the consequences for AI?

The development of language is intertwined with the histories of people. Many of those who experienced colonialism and empire have seen their own languages being marginalised and not developed to the same extent as colonial languages. African languages are not as often recorded, including on the internet.

So there isn’t enough high-quality, digitised text and speech to train and evaluate robust AI models. That scarcity is the result of decades of policy choices that privilege colonial languages in schools, media and government.

---

Read more: AI chatbots can boost public health in Africa – why language inclusion matters

---

Language data is just one of the things that’s missing. Do we have dictionaries, terminologies, glossaries? Basic tools are few and many other issues raise the cost of building datasets. These include African language keyboards, fonts, spell-checkers, tokenisers (which break text into smaller pieces so a language model can understand it), orthographic variation (differences in how words are spelled across regions), tone marking and rich dialect diversity.

The result is AI that performs poorly and sometimes unsafely: mistranslations, poor transcription, and systems that barely understand African languages.

In practice this denies many Africans access – in their own languages – to global news, educational materials, healthcare information, and the productivity gains AI can deliver.

When a language isn’t in the data, its speakers aren’t in the product, and AI cannot be safe, useful or fair for them. They end up missing the necessary language technology tools that could support service delivery. This marginalises millions of people and increases the technology divide.

What is your project doing about it – and how?

Our main objective is to collect speech data for automatic speech recognition (ASR). ASR is an important tool for languages that are largely spoken. This technology converts spoken language into written text.

The bigger ambition of our project is to explore how data for ASR is collected and how much of it is needed to create ASR tools. We aim to share our experiences across different geographic regions.

The data we collect is diverse by design: spontaneous and read speech; in various domains – everyday conversations, healthcare, financial inclusion and agriculture. We are collecting data from people of diverse ages, gender and educational backgrounds.

Every recording is collected with informed consent, fair compensation and clear data-rights terms. We transcribe with language-specific guidelines and a large range of other technical checks.

In Kenya, through Maseno Centre for Applied AI, we are collecting voice data for five languages. We’re capturing the three main language groups Nilotic (Dholuo, Maasai and Kalenjin) as well as Cushitic (Somali) and Bantu (Kikuyu).

---

Read more: What do Nigerian children think about computers? Our study found out

---

Through Data Science Nigeria, we are collecting speech in five widely spoken languages – Bambara, Hausa, Igbo, Nigerian Pidgin and Yoruba. The dataset aims to accurately reflect authentic language use within these communities.

In South Africa, working through the Data Science for Social Impact lab and its collaborators, we have been recording seven South African languages. The aim is to reflect the country’s rich linguistic diversity: isiZulu, isiXhosa, Sesotho, Sepedi, Setswana, isiNdebele and Tshivenda.

Importantly, this work does not happen in isolation. We are building on the momentum and ideas from the Masakhane Research Foundation network, Lelapa AI, Mozilla Common Voice, EqualyzAI, and many other organisations and individuals who have been pioneering African language models, data and tooling.

Each project strengthens the others, and together they form a growing ecosystem committed to making African languages visible and usable in the age of AI.

How can this be put to use?

The data and models will be useful for captioning local-language media; voice assistants for agriculture and health; call-centre and support in the languages. The data will also be archived for cultural preservation.

---

Read more: Hype and western values are shaping AI reporting in Africa: what needs to change

---

Larger, balanced, publicly available African language datasets will allow us to connect text and speech resources. Models will not just be experimental, but useful in chatbots, education tools and local service delivery. The opportunity is there to go beyond datasets into ecosystems of tools (spell-checkers, dictionaries, translation systems, summarisation engines) that make African languages a living presence in digital spaces.

In short, we are pairing ethically collected, high-quality speech at scale with models. The aim is for people to be able to speak naturally, be understood accurately, and access AI in the languages they live their lives in.

What happens next for the project?

This project only collected voice data for certain languages. What of the remaining languages? What of other tools like machine translation or grammar checkers?

We will continue to work on multiple languages, ensuring that we build data and models that reflect how Africans use their languages. We prioritise building smaller language models that are both energy efficient and accurate for the African context.

The challenge now is integration: making these pieces work together so that African languages are not just represented in isolated demos, but in real-world platforms.

One of the lessons from this project, and others like it, is that collecting data is only step one. What matters is making sure that the data is benchmarked, reusable, and linked to communities of practice. For us, the “next” is to ensure that the ASR benchmarks we build can connect with other ongoing African efforts.

---

Read more: Does AI pose an existential risk? We asked 5 experts

---

We also need to ensure sustainability: that students, researchers, and innovators have continued access to compute (computer resources and processing power), training materials and licensing frameworks (Like NOODL or Esethu). The long-term vision is to enable choice: so that a farmer, a teacher, or a local business can use AI in isiZulu, Hausa, or Kikuyu, not just in English or French.

If we succeed, built-in AI in African languages won’t just be catching up. It will be setting new standards for inclusive, responsible AI worldwide.

– African languages for AI: the project that’s gathering a huge new dataset
– https://theconversation.com/african-languages-for-ai-the-project-thats-gathering-a-huge-new-dataset-266371