We welcome our first non-human guest to the podcast, the language model ChatGPT. The chatbot joins us to answer questions about science and faith. ChatGPT draws on extensive digital information to answer the same kinds of questions we have been asking our human guests over the past few years. The results are…interesting. Jim and Colin reconvene afterward to talk about what to take from it all.
@Dale will like this one, the robot knows about providence.
5 Likes
Now it also understands a key distinction with the cosmological argument (my text to ChatGPT is in bold):
In a situation where the number of steps can proceed to infinity, it would then be impossible for the number of steps to become actually infinite
Yes, that is correct. In a situation where the number of steps can proceed to infinity, it is still not possible for the number of steps to become actually infinite. This is because infinity is not a number that can be reached or attained through a finite process.
Instead, when we talk about an infinite process or an infinite sequence of steps, we are referring to a process or sequence that can be continued indefinitely without ever reaching a final state or end point.
For example, some iterative algorithms involve a process that repeats a certain set of steps over and over again, potentially an infinite number of times. However, even if this process continues indefinitely, it still does not reach an actual infinite number of steps. Instead, it simply continues without ever reaching a final result.
In summary, even if a process or sequence of steps can proceed to infinity, it never actually reaches infinity, but rather continues indefinitely.
Yes, I also got a kick out of the chatGPT recognising my particular Evolutionary Creation nuance around providence too. This is another fascinating Podcast.
2 Likes
After 20 years of talking about this, I can honestly say, I could not have said it better myself
I’m looking forward to this. However, I remember someone from the YEC point of view in the late '80s saying that they fed information into a computer and came up with the conclusion that God was more likely to have created, than evolution. I talked about that with a man (who I realized later was a Christian geologist, old earth and likely evolutionist, from Abbotsford BC @klw ), and he quietly responded, “Garbage in, garbage out!” That, in retrospect, seems profound to me.
Thanks.
4 Likes
Interesting podcast. The letter of appreciation at the end was a little disturbing as it sounded like every letter of appreciation that has been written. That is turn tells us that perhaps we are not as unique and special as we think we are, but rather more like AI than we want to admit.
4 Likes
Any chance that you’ll finally “give it a rest”? 
2 Likes
It was a fun episode. I also liked the earlier lent one though I don’t practice lent. I enjoyed the dialogue though at times I was wondering about some things but I think Jim cleared it all by explaining that EC/TE does not necessarily mean god was somehow making a evolution work through unexplainable supernatural miracles.
I use to prefer the term theistic evolution but now I use evolutionary creationism and I use it mostly because I like annoy those who use it for young earth / old earth ID. It’s also why I now will often refer to young earth creationist as baby world theists xd.
1 Like
On my days when I am not as nice, I use the term “young earthists” when conversing with those who use terms like Darwinist and Evolutionist as pejoratives just to keep the playing field level.
3 Likes
Because ChatGPT has the whole internet at its fingertips…
(One thing that appears to be a common misconception is that it is online with access to the internet. We access it through the internet of course, but it does not ‘research’ online for its answers.)
Here, I’ll ask it:
“…curated and selected” is definitely of note.
(If you have deduced that I’m just starting the podcast, you would be correct. ; - )
1 Like
We did not change any words or add any words or even change the order of anything, but did remove some repetitions, of which there were many.
Ain’t that the truth! It would not pass the Turing test!
I don’t recall where I saw this, but it was an interesting video or article that looked at the factors that affect the perception of intelligence or personhood.
If a wall is introduced where you don’t know whether you are talking with a human being or a machine, how long will it be before the ability to tell the difference is a real issue?
We are approaching that threshold. 20 years ago this was not a possibility in internet chatrooms, if it isn’t today, it soon will be.
I enjoyed the podcast!
I must say I was really impressed with ChatGPT’s ability to organize and express information in the English language. (Aside: are there versions of ChatGPT for other languages?) I can understand the sentiment “I couldn’t have said it better myself”.
One thing that didn’t quite come out in the interview, since it wasn’t in real time, is: what were the response times of ChatGPT? I would be even more impressed if it gave those answers with no discernible time delay. (I suppose in some cases no time delay would be an indication that it wasn’t a person).
ChatGPT is very open (can I say honest?) that it’s an AI language model, and that it doesn’t have opinions. I commend the creators of ChatGPT for that.
I agreed with much of what Colin and Jim said in their concluding remarks. (I laughed out loud when they talked about what it means to understand the meaning of language 
). They mentioned the word “emotion”. It seems like great works of creativity by humans always come from deep emotions, or longings of the soul. I don’t think AI is anywhere close to being creative in that sense. For example, could AI make a speech “I have a dream that one day this nation will rise up and live out the true meaning of its creed…” – Martin Luther King Jr.
Or, could it write the words “It was the best of times, it was the worst of times, it was the age of wisdom, it was the age of foolishness, …”? – Charles Dickens, A Tale Of Two Cities
In terms of scientific discovery, could AI come up with a truly revolutionary discovery like Wallace and Darwin did with regard to the Origin of Species by Means of Natural Selection?
My predisposition is that we are not anywhere near endowing AI with such creativity, but I wonder if there is any reason why, in principle, we couldn’t understand enough about the human brain to do that.
A couple of us poked around the program a few weeks ago in a thread here. One person had it write a story about Jesus convincing Santa Clause why he shouldn’t kill a vampire child. It was a believable story with a moral and some character development. I had it write a pirate song about a bad day at work and realized it will write a different song each time you ask for one. It does not save its responses and creates a new response each time. This is why it makes plagiarism undetectable. Most of its responses were created in as much time as internet search results from 25 years ago. If it were matched up with a voice synthesizer it would seem nearly realistic. Customer service call centers will be like Blockbuster video stores in a few years.
I have used it personally for the first time last night, and as soon as I pressed send it started typing. You can see each letter that it types, just like you could yourself on a computer. it’s very wild.
3 Likes
Interesting that it brought up theistic evolution immediately as an example of integration between science and faith. I don’t what else I could have expected, especially in the West.
@LM77 – Yes, the distinction between TE and EC seemed pretty apt.
Providence:
God’s activity is seen as working through the laws of nature and the natural processes that God has set in motion. This way of understanding God’s activity in the natural world emphasizes God’s sovereignty and wisdom, rather than a need for constant miraculous interventions.
Yay. 
1 Like
That would be ‘it’ passing the Touring test.
The thing about perception is that some people could fail the test
1 Like