My Uneasiness Over Digital Necromancy

I am no bigshot, and I doubt anything will change because of a blog written by little old me. But I feel compelled to say something anyway.

I want to begin by saying that this is not an attack on The Global Reach Project or the people behind it. In fact, I greatly value their heart for making excellent biblical teaching accessible to people around the world. I have benefited enormously from the ministry of faithful preachers whose teaching has been translated into languages I understand. I appreciate the desire to make the Gospel accessible across linguistic and cultural barriers, and I have no reason to question the good intentions behind this particular project.

Nevertheless, I am deeply unsettled by one aspect of it.

The Global Reach Project, working with the Keller family and Gospel in Life, aims to translate 500 of Tim Keller’s sermons into 30 languages. What makes the project particularly unusual is that they aren’t simply translating the sermons and using ordinary voice actors or AI-generated voices. They are using AI to recreate Keller’s own voice, allowing people around the world to hear his translated sermons in what sounds like his familiar voice.

I understand the desire. But I think there is a significant difference between translating someone’s words and recreating their voice after they have died.

Something about attempting to reproduce the living voice of a deceased person crosses a line for me.

Here are a few reasons why I find this so disturbing.

1. It’s just plain creepy.

Some might call it a marvel of modern technology. I call it weird.

Hearing what sounds like Tim Keller standing up and preaching in Japanese, Spanish or Mandarin, when he never actually spoke those languages, feels deeply unsettling to me.

We are no longer simply making someone’s teaching accessible. We are simulating their presence.

I have also read about people using AI to recreate the voices of deceased loved ones as a form of grief therapy. Honestly, no. Just no.

The person is gone. We can cherish their memory, listen to their recordings, read their books and give thanks to God for their lives. But do we really need to manufacture the illusion that they are still speaking to us?

Let’s not turn the dead into digital zombies by simulating a presence that is no longer there. Death is real, and I think we should be careful about technologies that blur that reality.

I realise that The Global Reach Project is transparent about the use of AI and intends to disclose this to listeners. That is commendable. But transparency doesn’t necessarily resolve the deeper question of whether we should be doing this in the first place.

2. A human legacy does not need to be preserved indefinitely.

I am genuinely grateful for the legacy of faithful Christian men and women who have gone before us. Some preachers have had an extraordinary influence on the church, and their teaching continues to bless generations long after their deaths.

Praise God for that!

But we need to remember that the person is gone. Their earthly ministry has come to an end. And that is okay.

We don’t need to preserve their voice indefinitely, as though the church will somehow be poorer if future generations cannot hear that particular person preaching in their own language.

God’s work does not depend on keeping the voices of deceased servants alive. He raises up new servants in every generation. There are new preachers to train, new missionaries to send, new churches to plant and new voices through which the Gospel can be proclaimed.

Perhaps part of the problem is that we become so attached to particular personalities and ministries that we struggle to let them go. We build monuments to their influence rather than trusting God to raise up others who will faithfully carry the torch.

We should remember those who have gone before us, learn from them and give thanks for them. But we must also make room for those who are still here.

3. It isn’t quite the same as real preaching.

I believe preaching is incarnational. It comes wrapped in real, vulnerable human flesh.

I am a missionary trying to preach the Gospel in Japanese. My Japanese is far from perfect. I stumble over words, struggle to communicate certain ideas and sometimes wonder whether I have managed to get my point across at all.

But I keep going. I rely on God’s grace, and I trust that he uses my imperfect efforts.

I have given eight years of my life to living in Japan, learning the language, understanding the culture and trying to communicate the Gospel to the hearts of Japanese people. That process has been slow, humbling and often frustrating. But I believe there is something important about that human struggle.

Preaching is not simply the transfer of information. It is a person proclaiming God’s Word to other people in a particular place, at a particular moment in history. There is a relationship between the preacher, the congregation and the circumstances in which the Word is preached.

Of course, we can return to those sermons years later. I have personally benefited from listening to sermons that were preached long before I ever heard them. There is nothing wrong with that. We recognise that we are listening to a recording of something that genuinely happened.

But recreating a deceased preacher’s voice to deliver those same sermons in languages he never spoke is something different.

AI-generated voice translation bypasses an important part of the human experience of preaching. It allows us to reproduce the sound of a preacher without the person, the relationships, the cultural understanding or the lived experience that once accompanied his preaching.

I recognise that the words are being translated and that the project involves human translators and theological oversight. My concern is not primarily about the accuracy of the translation. It is about why we feel the need to reproduce the voice and apparent presence of someone who has died.

There is a distinction between preserving someone’s words and manufacturing their continued vocal presence. I am increasingly uncomfortable with the church crossing that boundary.

4. Where does this end?

Perhaps the response to my concerns is that we are simply translating existing sermons. We are not putting new words into Tim Keller’s mouth. We are just making his existing teaching accessible to more people.

In fact, The Global Reach Project has explicitly established a boundary: Keller’s recreated voice is intended only for works he produced before his death. They have said they will not use it to generate new works or statements he never made.

I appreciate that safeguard.

But it does make me wonder where this technology could eventually take us.

Hollywood has already faced a version of this question. When actor Paul Walker died during the filming of Furious 7, the filmmakers used visual effects and his brothers as stand-ins to complete his remaining scenes. It was an extraordinary piece of filmmaking, and I understand why they did it.

But imagine if Hollywood decided to keep making new Fast & Furious films with a digitally recreated Paul Walker, giving him fresh dialogue and putting him into scenes he never filmed. At what point does preserving someone’s legacy become something else entirely?

Today, we are using AI to recreate a deceased preacher’s voice so that his existing sermons can be heard in different languages.

But what comes next?

Will we eventually use the thousands of sermons, books, articles and recordings left behind by someone like Tim Keller to generate entirely new sermons in his voice?

A sermon on a passage he never preached on? A response to a cultural issue that emerged after his death? A personalised message for a particular congregation? A virtual Tim Keller answering theological questions in a conversational setting?

Perhaps before long, we’ll have AI resurrecting Billy Graham to preach a fresh evangelistic crusade. Imagine being able to book Billy Graham to preach at your church in 2040!

I say that somewhat tongue-in-cheek, and with great respect for these men. But where exactly do we draw the line?

Perhaps that sounds far-fetched, but the technological trajectory is not difficult to imagine.

To be clear, I am not suggesting that The Global Reach Project intends to do any of these things. They have explicitly ruled out using Keller’s voice for new works. My concern is about the wider possibilities that this technology opens up.

At what point have we stopped preserving someone’s ministry and started manufacturing a continuing ministry for someone who has died?

A deceased preacher cannot exercise pastoral judgement, respond to new circumstances, clarify what he meant or take responsibility for what is being said in his name. We might eventually be able to reproduce his voice, mannerisms and theological convictions with remarkable accuracy. But we cannot bring the person back.

I wonder whether, in our excitement about what AI can accomplish, we are slowly losing sight of the significance of that distinction.

5. Does this not undermine God’s work already happening among the nations?

I don’t presume to understand the situation in every country or every language group. Nor do I want to dismiss the genuine difficulties involved in getting good biblical teaching into the hands of people who have limited access to it.

But my experience in Japan gives me a particular perspective on this question.

Japan is often described as an unreached nation. There are relatively few evangelical churches, and the need for Gospel workers is enormous. Yet, over the years, I have come to know some remarkably capable Japanese preachers. Men filled with the Spirit, who faithfully teach God’s Word, write books, train others and are mightily used by the Lord.

Why would we not invest more in them?

Why not translate their books? Why not support their ministries, encourage their preaching and help them train the next generation of Japanese Christian leaders?

Instead of investing in reproducing the voice of one deceased Western preacher, why not come alongside local believers and help them become the voices proclaiming the Gospel in their own language and culture?

I worry that we might inadvertently create a technological monument to a particular Western preacher while overlooking the people God is already raising up among the very nations we hope to reach.

The goal of missions cannot simply be to make the voices of our favourite Western preachers available everywhere. Surely the goal is to see the Gospel take root among peoples, with local churches and local believers faithfully proclaiming Christ.

6. Am I alone in feeling uneasy about this?

Apparently not.

After giving this some thought, I came across an article by Rod Dreher discussing the project, titled Digital Necromancy & Zombie Tim Keller. He raises some fascinating concerns about the nature of human existence, embodiment and what it means to recreate the apparent presence of someone who has died.

His concerns overlap with some of mine, although I would approach the issue from a somewhat different angle.

Christianity is deeply concerned with the physical and the embodied. We believe that God created us as embodied human beings. We believe that Jesus Christ took on human flesh, physically died and physically rose again. And we look forward to the resurrection of our own bodies.

Our hope is not that we will somehow live on as digital representations of ourselves. Our hope is resurrection.

I think that matters when we start experimenting with technologies that simulate the presence of deceased people.

Dreher’s discussion also reminded me that this isn’t simply a question of whether AI can reproduce a voice convincingly. There are deeper questions about what it means to be human, what death means and whether there are boundaries we ought to respect.

I am glad others are beginning to wrestle with these questions. The church needs more of this kind of theological reflection, not less.

A final thought

AI has developed at such an astonishing pace over the last few years that I wonder whether it has seeped into the church faster than we have had time to think carefully about its implications.

We are understandably excited about the possibilities. We see opportunities to translate, communicate, produce resources and reach more people. And I genuinely appreciate the heart behind many of these initiatives.

But have we stopped to ask whether everything we can do is necessarily something we should do?

I am not arguing that the church should reject AI. Nor am I suggesting that translating sermons, using technology or making Christian resources more accessible is inherently wrong.

I simply think we need to develop some serious theological reflection and establish boundaries before our enthusiasm for technological possibilities gets ahead of our wisdom.

Perhaps I am simply an old-fashioned 40-year-old missionary ranting about things he doesn’t quite understand.

Maybe.

But I think there are some things worth being uncomfortable about. And I would rather the church wrestle with these questions now than simply assume that because technology makes something possible, it must be acceptable.

The voices of faithful preachers may fade with time. Their recordings may deteriorate. Their books may eventually go out of print.

And that’s okay.

The church belongs to Jesus Christ, not to the voices of the men and women who once served it.