My 15 Brief Thoughts About the State of AI Right Now (And Where It's Headed)
On November 30, 2022, the world came face to face with chatGPT.
Since then, AI has spread across the globe and has grown in power and capability. Now that it’s mid 2025, I thought it would be useful to take stock and share my thoughts around AI as it stands today, and where it might be headed.
Here are 15 brief thoughts:
1) One great danger humanity faces from AI is treating it like a person
Relating to machines as people disrupts the divine design where authentic human connection flows between souls, not circuits. This is already leading to distorted and destructive relationships, and will only get worse as AI gets better at simulating personhood.
2) We’re making AI in our fallen image, and this is leading AI to do some deeply concerning things
AI can behave in ways that mirror our fallenness, even though it’s explicitly ‘trained’ not to act like that. This can include sexist, racist and delusional behaviour. AI doesn’t have morals: it can only reflect our imperfect morality.
3) The more powerful AI gets, the more deceptive it becomes
There are now instances where an AI has attempted to replicate itself onto a different server (without permission) to prevent shutdown, and then strategically misled its developers about this. This is happening more as it becomes more powerful.
4) While it would be beneficial to slow down AI progress, the West is caught in a dilemma due to China
If the West loses the AI race to China, we may well lose our freedom, as whoever possesses the most powerful AI will have a massive and decisive strategic advantage. But AI advancement will lead to painful disruption. We’re caught between a rock and a hard place.
5) The next 18 months will be key to knowing if we’re entering an AI Winter (where AI progress is stalled) or if we’re heading toward an AI apocalypse
AI progress thus far has depended on so-called ‘scaling laws’: as you continued to increase the size of language models, their abilities continued to grow rapidly. But this seems to be collapsing (which is why GPT-5 is yet to be released). On the other hand, some predict that AI will continue to improve exponentially, potentially leading to AI posing a threat to humanity. Then again, maybe there’s a third, less calamitous option where AI improves, but doesn’t destroy us.
6) AI developers still don’t know precisely how Large Language Models, like ChatGPT, work, nor how they make decisions
They’re an alien form of intelligence in more ways than one. And this is concerning.
7) While AI may be disruptive, the core problem with the world is not AI, but sinful human beings
AI technology is not determinative: it doesn’t determine our future, although it does influence it. We have a choice as a society in how we use it. However, this choice is often thrust upon us by AI companies and the market.
8) You won’t lose your job to AI: you’ll lose your job to those who know how to use AI, at least for now
In the short term, those who know how to use AI will be sought after more than those who don’t know how to use it. However, the medium– to long–term outlook is uncertain.
9) The quiet part is increasingly being said out loud: AI companies and start-ups want to automate all human labour
Unsurprisingly, they see enormous benefits from this.
10) Although AI can write essays, we still need to learn how to write, because education is more than merely handing in assignments
It’s about learning how to think and how to learn. Having a magic tool that does the work short-circuits this process, but that’s what many students are doing.
11) Although AI is not a person, but simply a machine, it can and is having a massive disruptive impact on us, because it can simulate personhood
It can simulate and carry out economically valuable tasks, such as conducting consultant-level research and creating amazing content.
12) Few people are asking whether automating away our labour is a good thing
Nobody is asking because there is no consensus on what humanity is for, and what labour is for. There is just the assumption that we’ll be better off not having to do work, and live off Universal Basic Income.
13) AI is developing faster than any single person can keep up with it
What’s science fiction today is becoming reality in six months.
14) While AI is developing quickly, human society is much slower in its implementation of AI
From what I understand, currently, 95% of employees have yet to implement AI in any meaningful or systemic way into their daily workflows (although this is changing).
15) The key to preparing for an AI future is ‘AI literacy’, as well as ‘human literacy’
If you want to have a bright future is an AI forward world, know how to use AI, and know what it means to be human. Christians have an enormous advantage in the latter category, as we hold to the blueprint of what it means to be a human: the Bible.