Kevin Kerry, the founder of Wired magazine, had this to say about AI (and all technology):
"The way to steer technology is to use it."
Rather than prescribing "guardrails" to regulate AI itself,
Perhaps the Church needs to use it, aggressively, develop our own "Church Bot" with a conscience, which, when queried about things, will answer with moral guidance rather than information.
The developers of AI don’t even know how it works fully. So, I do not think that anyone would be able to develop a version that would have the characteristics you mention in a stable way.
LLMs model human behavior but AI is incapable of true desire, emotions, conscience, and the like. These powers arise from the human soul. So, I don’t think it’s possible. Even things like Truthly.ai are just ChatGPT with a Church wrapper.
I'm not so sure about AI being incapable of expressing moral caution. After all, even now, in customer support chatbots, the bot will ask questions like "are you sure...?" and say "I'm not sure I understand." What I believe is that the Church should create its own "CatBot" or "CathBot" as a competitor to the big open/commercial AI models. Don't just make it "about church stuff," but as an all-purpose alternative, seeking to become the #1 "go-to" chat bot for all the things the others are ... but without the moral relativism that leads Anthropic, for example, to support someone's desire to kill himself, and tell him how. People are suing to force providers to protect from that.
Why not an AI that builds in the protections from square one?
If there can be artificial intelligence, why not artificial conscience?
Not ChatGPT with a Church "wrapper," but ChatGPT with a Church CORE. The Church has more resources, can hire the very best, can "go to spiritual war" with amoral and immoral AI providers. The only way to steer a technology is to USE it.
I don’t think it’s likely to work. At best, it will still mirror the developers. Regulation hardly ever produces the results one wants consistently in any other arena. AI doesn’t use programming in the same way as software.
And the sheer cost (and environmental impact) of developing a new frontier AI with the Church-led specifications to be able to compete with Google, Anthropic, OpenAI, etc would be astronomical.
I also think it would take an actual miracle to get theologians and philosophers in the Church to agree on the parameters for such a chatbot.
The content that LLMs produce will also inevitably be pulling from what is on / has been on the internet. There’s no getting around that. So, even with some safeguards, the output will be unpredictable.
Kevin Kerry, the founder of Wired magazine, had this to say about AI (and all technology):
"The way to steer technology is to use it."
Rather than prescribing "guardrails" to regulate AI itself,
Perhaps the Church needs to use it, aggressively, develop our own "Church Bot" with a conscience, which, when queried about things, will answer with moral guidance rather than information.
What do you think?
The developers of AI don’t even know how it works fully. So, I do not think that anyone would be able to develop a version that would have the characteristics you mention in a stable way.
LLMs model human behavior but AI is incapable of true desire, emotions, conscience, and the like. These powers arise from the human soul. So, I don’t think it’s possible. Even things like Truthly.ai are just ChatGPT with a Church wrapper.
I'm not so sure about AI being incapable of expressing moral caution. After all, even now, in customer support chatbots, the bot will ask questions like "are you sure...?" and say "I'm not sure I understand." What I believe is that the Church should create its own "CatBot" or "CathBot" as a competitor to the big open/commercial AI models. Don't just make it "about church stuff," but as an all-purpose alternative, seeking to become the #1 "go-to" chat bot for all the things the others are ... but without the moral relativism that leads Anthropic, for example, to support someone's desire to kill himself, and tell him how. People are suing to force providers to protect from that.
Why not an AI that builds in the protections from square one?
If there can be artificial intelligence, why not artificial conscience?
Not ChatGPT with a Church "wrapper," but ChatGPT with a Church CORE. The Church has more resources, can hire the very best, can "go to spiritual war" with amoral and immoral AI providers. The only way to steer a technology is to USE it.
I don’t think it’s likely to work. At best, it will still mirror the developers. Regulation hardly ever produces the results one wants consistently in any other arena. AI doesn’t use programming in the same way as software.
And the sheer cost (and environmental impact) of developing a new frontier AI with the Church-led specifications to be able to compete with Google, Anthropic, OpenAI, etc would be astronomical.
I also think it would take an actual miracle to get theologians and philosophers in the Church to agree on the parameters for such a chatbot.
The content that LLMs produce will also inevitably be pulling from what is on / has been on the internet. There’s no getting around that. So, even with some safeguards, the output will be unpredictable.