AI Investment to Reach $632 Billion by 2028, Shift Toward Human‑Centric Uses Urged

*TL;DR: Global AI spending will top $632 bn by 2028; experts say the next wave must deliver tangible human benefits.

Context The artificial‑intelligence market is entering a new financial phase. Capital inflows that once funded experimental models are now being measured against real‑world impact. Industry analysts warn that without a clear link to human needs, the surge in funding could become a costly bubble.

Key Facts - The International Data Corporation projects worldwide AI expenditure to reach $632 bn (about KES 83.4 trillion) by 2028. The forecast marks a pivot from novelty projects to applications that improve daily life. - PwC’s economic model estimates AI could add more than $15 trillion to global GDP by the end of the decade, but only if the technology penetrates legacy sectors such as healthcare, energy and agriculture. - Machine‑learning value is most evident in three areas: speeding drug discovery by analysing complex biological data, stabilising power grids through predictive load management, and forecasting weather to protect food supplies.

What It Means Investors and developers now face pressure to re‑engineer AI pipelines. Rather than building ever larger language models for entertainment, firms are urged to start with a problem—low‑bandwidth connectivity, unreliable electricity, or limited digital literacy—and design models that work within those constraints. In drug development, AI can cut years off the research cycle, potentially lowering costs and expanding access to new therapies. Energy utilities can use predictive algorithms to balance supply and demand, reducing blackouts and emissions. Agricultural AI that reads hyper‑local soil data can boost yields while limiting fertilizer use, provided the software runs on basic mobile devices common among smallholder farmers.

The shift also has geopolitical implications. African startups, familiar with intermittent infrastructure, are positioned to create AI solutions that respect local languages and data contexts, avoiding the pitfalls of imported models that ignore regional nuances. Successful deployment will require rigorous testing under real‑world conditions—handling dropped connections, corrupted data and erratic user input—to ensure safe failure modes.

As the $15 trillion economic upside hinges on these outcomes, the market will likely reward firms that demonstrate measurable human impact. The next reporting period will reveal whether AI spending translates into concrete improvements in health, energy reliability and food security.

Watch next: Early‑stage pilots that tie AI to drug pipelines, grid management and climate‑smart farming will be the barometer for the sector’s promised economic boost.