In this interview with PREMIUM TIMES on the sidelines of COP17 in Ulaanbaatar, Mongolia, Oluseyi Ifatimehin, a Professor of Geography (Environmental Resource Planning) at Kogi State University, Nigeria, and Science and Technology Correspondent to the United Nations Convention to Combat Desertification (UNCCD), outlines the scientific and institutional gaps confronting Nigeria and other African countries and proposes stronger land and drought observatories, artificial intelligence, satellite technology, indigenous knowledge and improved science-policy mechanisms to tackle the challenges.

Excerpt:

PT: What are the biggest scientific gaps preventing Nigeria and other African countries from effectively addressing land degradation and drought?

Ifatimehin: From my perspective, Africa does not primarily suffer from a lack of scientific knowledge. The bigger problems are gaps in data continuity, local-scale evidence, technology deployment and the capacity to translate evidence into decisions. For Nigeria, one major gap is the lack of sufficiently dense, continuous and interoperable datasets on soil health, land degradation, groundwater, vegetation, drought indicators and land-use change. Satellite data are increasingly available, but they need to be combined with reliable ground observations to produce information that is useful at the farm, watershed and community scales.

A second gap is predictive capacity. We need to move from asking, “Where has degradation or drought occurred?” to ask, “Where is it likely to occur next, when will it happen, how severe will it be, and what should farmers and governments do now?”

A third gap is the shortage of sustained investment in African scientific institutions, laboratories, field-monitoring networks, data infrastructure and technical personnel. Nigeria’s priority should therefore be to establish stronger national and regional land and drought observatories, integrate satellite and ground data, strengthen universities and research institutions, and develop indicators that measure outcomes. These could include hectares of degraded land restored, changes in soil organic carbon, vegetation recovery, water availability and the number of farmers receiving actionable early warnings.

PT: Why is there a disconnect between research institutions and policymakers, and how can COP17 help close that gap?

Ifatimehin: The disconnect is partly because research and policymaking often operate on different timelines and communicate in different languages. Researchers may produce excellent scientific papers, but policymakers need concise answers to practical questions: What is the problem? What will it cost? What intervention works? Where should it be implemented? What result can we expect within three or five years? There is also insufficient institutional linkage between universities, government ministries, agencies and communities. Research funding is frequently project-based, while policy implementation requires long-term evidence and monitoring. COP17 can help by strengthening the UNCCD Science-Policy Interface (SPI) and encouraging countries to establish stronger national science-policy mechanisms. Universities should not be viewed simply as institutions that produce publications; they should become knowledge partners in national land-restoration programmes.

In line with this, proposals from two Nigerian universities — the Federal University Lokoja (FUL) and Confluence University of Science and Technology (CUSTECH), Osara — were presented at the COP17 side events.

FUL presented two proposals: ECO-HUB for Land Restoration Technical Education, focused on ecopreneurship, restored land, resilient food systems and shared prosperity; and Building a University-led Knowledge Curriculum and Research Platforms for Sustainable Rangeland Resource Governance, Livestock Reform and Peace-building for Nigeria’s Sub-humid and Humid Ecological Zones.

CUSTECH presented Measuring Nature, Building Skills and Restoring Land: Bio-resource Quantification, Ecosystem Evaluation and Land Degradation Neutrality Capacity in Kogi State, Nigeria.

For Nigeria, I would like to see a measurable mechanism whereby scientific baselines, measurable targets, independent monitoring and periodic evidence reviews accompany major land-restoration policies and programmes. The ultimate test is simple: Can scientific evidence change a government decision, improve an investment or change what happens on the farm? If the answer is yes, then we are closing the science-policy gap.

PT: What technologies could Nigeria realistically deploy to predict drought and prevent its worst impacts?

Ifatimehin: Nigeria does not need to wait for futuristic technology. Much of the technology required already exists; the challenge is integrating it and deploying it at scale.

First is satellite Earth observation. Satellite systems can monitor vegetation stress, soil moisture, rainfall anomalies, surface water and land-use change across Nigeria. This allows the government to identify emerging drought conditions over large areas much faster than conventional field surveys.

Second is artificial intelligence and machine learning. AI can combine satellite observations with rainfall, temperature, soil, crop, hydrological and historical drought data to identify patterns and generate localised drought-risk forecasts. Third is multi-hazard early-warning systems. Nigeria should connect meteorological and hydrological forecasts with agricultural advisory services so that a warning becomes an actionable message — for example, when to plant, which crop or variety is more appropriate, when to conserve water, or where livestock movement and water resources may become constrained.

Fourth is climate modelling and seasonal forecasting, which can support planning several months ahead. But technology alone is insufficient. Nigeria needs an end-to-end drought information system linking observation to forecasting, early warning, communication, early action and impact assessment. A measurable target could be to ensure that drought warnings reach vulnerable farming and pastoral communities before critical agricultural decisions are made, rather than after losses have already occurred.

PT: How can modern science and Indigenous knowledge be combined at the community level?

Ifatimehin: The answer is not to choose between modern science and Indigenous knowledge. We need to put them into dialogue. Farmers and pastoralists possess generations of knowledge about rainfall patterns, soil characteristics, pasture availability, water sources, drought indicators, livestock behaviour and locally adapted crops. Modern science brings remote sensing, climate modelling, soil analysis, forecasting and new technologies.

The most effective approach is therefore co-production of knowledge. Scientists should work with communities to test Indigenous observations against meteorological, ecological and satellite data and determine what is reliable, where it works and under what conditions. For example, if pastoral communities have traditional indicators for anticipating pasture or water scarcity, these can be assessed alongside satellite vegetation indices, rainfall forecasts and groundwater information. Where both sources provide consistent signals, the resulting early-warning system is likely to have greater local credibility and uptake.

We should also ensure that communities are not merely treated as recipients of scientific information but as knowledge partners.

ALSO READ: COP17 advances $1.3bn for land restoration, puts rangelands at centre

A measurable approach would be to establish community demonstration sites where restoration interventions are jointly designed, tested and monitored, with indicators such as vegetation recovery, soil health, water availability, livestock productivity and household livelihoods.

PT: What should governments, universities and development partners do differently to move African innovations from laboratories into practical solutions?

Ifatimehin: Africa needs to change the way it measures research success. A scientific publication is important, but it should not be the final destination.

Governments should create innovation-to-implementation pathways in which promising technologies are identified, field-tested, independently evaluated and then incorporated into national programmes and procurement systems.

Universities should strengthen technology-transfer offices, innovation hubs, demonstration farms and partnerships with farmers, pastoralists, government and industry. Research funding should include resources for field validation, scaling and adoption, not only laboratory research.

Development partners should move beyond financing short-term pilot projects. Where a technology has demonstrated effectiveness, financing should support the transition from pilot to demonstration, scale-up and institutional adoption.

The private sector is also critical because governments and donors alone cannot finance the scale required.

For Nigeria, I would propose a Land Restoration Innovation Pipeline with clear stages: Research → Field testing → Independent validation → Community adoption → Investment readiness → Scale-up → Impact measurement. Success should be measured not simply by the number of papers or projects produced, but by outcomes such as hectares restored, farmers and pastoralists reached, increases in soil and water productivity, drought losses avoided, jobs created, technologies commercialised and sustained improvements in livelihoods.

My central message from COP17 is that Africa does not need to reinvent science; it needs to connect the science we already have to decisions, finance and action.

Nigeria has universities, research institutions, satellite technologies, climate scientists, Indigenous knowledge and a growing innovation ecosystem. The opportunity now is to connect these assets through stronger science-policy institutions, digital knowledge platforms, predictable financing and measurable implementation. “Restoring Land, Restoring Hope” must therefore mean moving from knowledge to action, and from isolated projects to nationally and locally scalable solutions.

PT: Thank you so much for your time, Prof.

Ifatimehin: Thank you for the opportunity. Bye.

At the Hot Chips conference on Tuesday, OpenAI shared a more detailed look at Jalapeño, including the first batch of benchmark results for the new system. Tested on Semianalysis’s InferenceX benchmark, Jalapeño registered both more tokens per user and more throughput per kilowatt than the currently available state-of-the-art inference processors.

“The bottom line is that the results show a very, very significant performance advance over state of the art,” said Richard Ho, OpenAI’s head of hardware, in a press call. “Jalapeño can serve more AI work per unit of power, while also returning responses more quickly. It’s very efficient to serve a lot of customers, but it can also be very low latency.”

Notably, that comparison is against an Nvidia Blackwell system — but by the time Jalapeño reaches full deployment, the competition may have advanced significantly. Ho estimated that Jalapeño would deploy at the end of 2026 “in very small volumes,” with more significant deployment coming in 2027.

First announced last October, Jalapeño was developed by OpenAI in close collaboration with Broadcom, with OpenAI’s own models assisting in the development process. The company plans to make Jalapeño a multigenerational platform, allowing AI products, models, chips and memory all developed in concert.

Because of that full-stack approach, OpenAI was able to address specific phases in the inference process that often cause friction during inference processing. In particular, Jalapeño is designed to minimize delays during the prefill and communication phases of processing, which OpenAI says often act as bottlenecks.

“We designed Jalapeño to minimize data movement and communication delays,” the company said in a blog post presenting the results. “This means that model state, including the KV cache used while generating a response, can be explicitly placed and kept local while the system activates the right combination of compute, memory, and networking for each inference phase.”