We have always produced machines to reduce our labour. These machines are necessarily stronger than we are. The Industrial Revolution produced machines that could build bigger, better and faster than human hands could manage.
In the 20th century, we produced machines that could calculate more and faster than we could, machines that could store and remember more information than we could. We then produced machines that could combine this and beat us at chess. In other words, by one metric, even the computers of the 20th century were ‘smarter’ than us. Our software could reveal more in seconds than our library research could have produced after hours of page turning.
The last years represent another threshold. We have now produced machines that can mimic and understand language. Combined with massive storage, and a huge range of examples, we now have machines that can write faster, and sometimes, better than we can: essays, descriptions, instructions, poems, riddles, jokes, speeches, or sermons. Similarly AI can now code software better than we can, and perform visual design of graphics, websites, presentations and eventually, live-action movies, and simulations. Whereas we thought that creative activity would always be the domain of the human, we have now trained computers to mimic the creative process and perform it faster than we can.
For example, I asked ChatGPT to write a prayer by Charles Spurgeon. After about ten seconds, it had the following:
“O Lord, we would delight ourselves in You who are the chief of all delights. We thank You that of Your own good pleasure You bring salvation to sinful men. We confess that before You made us to differ, we were as others are, dead in trespasses and sins, having no hope and without God in the world. We bless You for the eternal covenant by which You foresaw us and chose us to be Your own, and for the precious blood of Christ by which we are redeemed. We pray that we may never lose sight of the exceeding sinfulness of sin, the infinite mercy of divine love, and the fullness of grace that dwells in Jesus. May the Holy Spirit who has begun the work of grace in us carry it onward to completion until we are complete in Christ, and find our joy and satisfaction in His service; for the Lord’s sake. Amen.”
Yes, for many years, Logos and Bibleworks had far more knowledge than any pastor could remember. But the knowledge was passive, so to speak. It was inert, waiting to to be probed by a thoughtful researcher. Now we have machines that can outdo the output of many a pastor, writing in a style that is (sadly) better than many pulpits.
In other words, just as we made machines that can lift more than we can and move faster than we can, we have now designed machines that can know more and communicate more in a shorter period of time. If thinking is defined as memory, factual recall, and the ability to swiftly convert factual recall into written (or verbal) communication, then we have designed machines that can think better than we can.
Economically, this will have similar effects to previous waves of automation. On the one hand, it will produce abundance. When time is saved, goods and services are produced faster. The abundance reduces prices, so a lot of things will become cheaper. AI is also likely to improve the quality of life in many respects. AI may design medicines, surgeries, software, machines, and technologies better than we ever could. It will have the computing power to test and discover chemical compounds with wondrous applications for medicine and human health.
Conversely, there will be massive job losses. Graphic designers, coders, and many other forms of communication and design work will soon be done faster, and often better, by AI. In fact, probably every realm of the economy will suffer loss: if AI can do your job better than you can, you’re out of a job. That includes doctors and lawyers. This job loss will produce massive social upheaval: we’re talking tens of millions of jobs. It is usually the case that jobs lost to automation soon find different jobs in the new economy, just as horse-shoe makers found work in car factories. This may be the case in an AI-economy, but it is not certain. In the interim, there will be the usual problems that go with job loss: depression, anger, despair, even social upheavals. At some point, unions will get involved and try to ban AI work.
And then there are the dystopian possibilities: defence networks controlled by AI, human freedoms eventually restricted by algorithms and enforced by restrictions on movement and trade, virtually unlimited powers of public and private surveillance. Beyond that, there is all that Lewis predicted in the The Abolition of Man: the dreams of the transhumanists, with neural links to AI information, and cybernetic links between the human brain and the Internet.
What does this look like for Christian ministry? In this series, I’ll suggest several implications.