DeepSeek, a Chinese-headquartered AI company, has been making headlines around the world. Unlike its American counterparts, which spent vast sums on developing their AI models, DeepSeek has achieved comparable results at significantly lower cost. There are lessons here for third sector organisations striving for high impact with constrained resources.

DeepSeek is a Large Language Model. Large Language Models predict the next word in a sequence, given the previous words in that sequence. Large Language Models themselves are remarkable. They have given computers the power to read and write at levels comparable to a human.

Large language models surpass humans on virtually all university-level written assessments - beating humans at written tests is now considered a solved problem. Brian Porter & Edouard Machery published a paper in November 2024 that showed AI-generated poetry is indistinguishable from human-written poetry and is rated more favorably.

Previous large language models have cost billions of dollars to build and train. DeepSeek shows that large language models can be built (relatively) inexpensively. ‘Relatively’ here still means millions of dollars, but that is a fraction of the cost of DeepSeek’s predecessors. The lower the cost of the underlying technology, and the more commoditized it becomes, the more cheaply it can be deployed and used by a wider range of businesses, charities, and individuals. Lower costs make it feasible for smaller organisations to experiment and innovate using the technology. Increased accessibility should lead to more personalised and niche applications, benefiting a wider spectrum of society.

DeepSeek’s success points to an important lesson: the value of AI will not lie in the technology itself but in how it is used. For non-profits, this will mean leveraging AI to enhance operations, extend reach, and improve service delivery.

Large Language Models are a general-purpose technology like electricity. Cheap, readily available electricity led to an explosion of technological innovation and industrial growth. Washing machines, refrigerators, vacuum cleaners, radios, and televisions - electricity is everywhere in the modern world. We don’t really think of electricity as a ‘technology’ anymore.

In the early days of electric lights, there were signs explaining that the lights should not be lit with a match, and that they were not harmful to health. This sounds comical to modern ears, but it highlights how new technologies often require a period of adjustment and new ways of understanding.

AI will be in everything over the next decade, as electricity is today. And, like electricity, we will no longer think of it as a ‘technology’. As artificial intelligence integrates into the fabric of daily life, it will become an invisible but indispensable part of our existence. AI will become so pervasive that its presence will be completely taken for granted. We no longer ponder the presence of electricity when we switch on a light.

The unfolding AI era will see a huge range of specialised applications tailored to specific tasks and industries. They will use LLMs as a toaster uses electricity. Each application requires a unique approach based on what you are trying to achieve. And just as you probably wouldn’t use an iron to toast bread, you probably shouldn’t use a chat-bot to review legal documents.

The charity sector is rightly renowned for its creative problem-solving methods, often operating under stringent budgets. It stands to gain significantly from the emergence of more accessible AI technologies. AI has the potential to be a transformative force, ushering in a new era of efficiency and effectiveness within the sector. Advances in AI offer a tangible pathway towards building a more inclusive and equitable society. A world where non-profit organisations can leverage technology to deliver more impactful solutions at a lower cost is a world where everyone benefits.

This article is sponsored by AutogenAI and written by Sean Williams. He is the founder and CEO of AutogenAI. He founded and was CEO of Corndel Ltd, and has worked in research, policy, business development and operational management for some of the largest, most successful public service providers in the UK.

---