ChatGPT

[R. Mansfield]

9 hours ago, Nathan Parker said:

I think that's a web version of the same search engine on their Desktop app (at least that's what it does for me). It's useful to have in the Accordance Web Browser though. 🙂

 

No, it's something completely new and not implemented into the desktop app yet. See more information here. 

[Mike Atnip]

1 hour ago, R. Mansfield said:

 

No, it's something completely new and not implemented into the desktop app yet. See more information here. 

I gave it a little spin and found that it does expand search terms to related words and concepts, but when it came down to it they were search terms that I knew (or should have known if I knew the topic!). So what it essentially does in incorporate several searches into one. Where it can excel is if the various search terms are interrelated and thus push hits to the top that are heavier in desired content. Where it may be weak is that if the hit has just one related search term (but contains very relevant information), it may be at the bottom.

It is interesting the Mark Barnes notes this: "We could of course, just pipe the query through to ChatGPT, but it simply hallucinates too often for serious trusted study, at least at the moment.  ... We've also tried "feeding" a large-language model with "answers" from our resources in order to get a natural language "answer" back. Using one of our private prototypes, I tried that, and it gave me a confident but wrong answer of 19 times to the same question. ... We're running lots of experiments and trying to follow the constantly-changing technology, but at the moment, we're not convinced that ChatGPT and its siblings are sufficiently accurate to incorporate into Logos."

AI-doped search engines can be useful, but they are still search engines that can only spit out what has been fed into them ... the good information along with the bad. When we realize that, we can use them wisely even in their limitations.

 

[Dr. Nathan Parker]

7 hours ago, R. Mansfield said:

No, it's something completely new

 

Thanks!

[docdave]

6 hours ago, Mike Atnip said:

We've also tried "feeding" a large-language model with "answers" from our resources in order to get a natural language "answer" back. Using one of our private prototypes, I tried that, and it gave me a confident but wrong answer of 19 times to the same question. ... We're running lots of experiments and trying to follow the constantly-changing technology, but at the moment, we're not convinced that ChatGPT and its siblings are sufficiently accurate to incorporate into Logos."

 

I hope that Accordance has been running experiments and brainstorming regularly about this too. I am quite certain that the latest ChatGPT is not ready for prime time commercial application, and given that so many are currently scrambling to identify all of its risks as well as what tasks it is actually suitable for,  I believe it was sort of foisted on the public in a irresponsible manner. Nevertheless, the technology will improve and, at the very least, large language models will soon have a profound effect on how all of us plan and do information search and retrieval. 

[William Cross]

7 hours ago, R. Mansfield said:

 

No, it's something completely new and not implemented into the desktop app yet. See more information here. 

Nice, but falls short of the AI ability to process natural language questions and sift through resources to compile a starting point of study. Really is just a compounded word search leaving way to many hits to sift through.

[Dr. Nathan Parker]

I've been playing around with ChatGPT for about a week for personal use (not particularly for any kind of work use, although my personal use of it is useful to learn more about AI). So far, I'm not seeing a benefit to paying $20/month for it for my own personal use, since my use of it would still be somewhat-casual. I have hit a couple of times when performance as weak due to heavy use, and I also maxed out the amount of queries in an hour with it, but I don't "rely" on it enough where when it's not functioning for me, it's OK.

 

For now, I seem to be using it for what I said above, a "search engine on steroids". I wouldn't use it for unethical purposes (I wouldn't use it to write a report, sermon, etc, and replace my own creative thinking). Where it has come in handy is after I've done my own reading and research, I've used it as a "cross checking" tool to help me synthesize the work I've already done. I've asked it to summarize books I've already read to ensure I see the "big picture" of what I've read, or I'll throw theological terms at it to summarize what I've already researched to ensure I'm not missing the "forest" for the "trees". I have found some inaccuracies from it (it gets some stuff wrong that my PhD dissertation will correct, but at least now I feel like my PhD dissertation is going to make a needed contribution!).

 

Other use cases would be for looking stuff up such as HTML coding, CSS coding, etc. Stuff like that was always a pain in the past to have to hunt down the coding for when searching with a conventional search engine. ChatGPT does a good job at giving me back content such as that and is a time saver there.

 

Overall, I've found it useful for certain use cases. It's not perfect. I wouldn't use it unethically, and I'm still not completely "wowed" by it (I just don't get excited about new tech anymore). I consider it an overall useful tool when used correctly and another time saver, but at the moment, I wouldn't pay $240/year for it until I could save more than that per year in productivity.

[Dr. Nathan Parker]

I haven't played with ChatGPT in a few days, and strangely, I'm not really missing it at all. After the "newness" wore off, I haven't really gone back to using it and still turning to a traditional search engine when I've needed to lookup something.

 

Two "concerns" I've had with ChatGPT in general are: 1. Who is "curating" the content, and are the curators truly giving us accurate information? 2. What is ChatGPT doing with the user data we're inputting in there? What kind of user privacy is there on it?

 

I'm not against AI in general, and I still see some place where Accordance could leverage/utilize AI. In terms of my own personal use of ChatGPT, I'm not sure now that the "newness" has worn off how much I'll personally use it, plus I still have outstanding concerns with it.

[Robb B]

On 7/14/2023 at 1:27 AM, Kristin said:

Am I the only one who doesn't like ChatGPT?

Nope. I don’t use it.

[cweber]

1 hour ago, Nathan Parker said:

Two "concerns" I've had with ChatGPT in general are: 1. Who is "curating" the content, and are the curators truly giving us accurate information? 2. What is ChatGPT doing with the user data we're inputting in there? What kind of user privacy is there on it?

 

I would counter that question with do you expect anything different from Google, who have dominated the search for decades) in this respect?

 

I get how people do not find generative AI helpful. You have to be focused in how you drive the tool. I use it daily for work and other programming interests because it is extremely useful for software development and capturing business requirements and processes. ChatGPT is one AI amongst many now days, but very beneficial. From considerable experience with these AI tools and what I see as an opportunity for improvement in Accordance, I remain convinced AI would be helpful for conversational search and assist with library management. This is especially true for those with a vast investment in library resources. It does not have to be more than that, yet would greatly uplift those two areas (depending on quantity of Texts and Tools) that I find wanting in Accordance.

 

[Dr. Nathan Parker]

I definitely see a use for generative AI, especially in your work case. It might be that after I go through my "cooling period" with it, that I find other use cases for it and start incorporating it into my workflow again. I'm still undecided on it. I tend to go through this with tech in general.

 

I could see where AI could definitely enhance "power searching" in Accordance and surface content in Accordance that's difficult to find through traditional searching. I was suggesting that Bible software look in these directions for years, even before "AI" became a household term. There are times when there is content in my vast library I'd like to surface and synthesize easily without using a complex search query. This is one area AI could mine those riches.

 

In terms of search engines, I never did use Google. I've gone back and forth with other ones that are supposed to be more privacy focused. Right now I'm giving Brave Search a try since it's supposed to be privacy focused and has its own independent search index. I'm still undecided if I'd use it long term. Before that, I was using Ecosia which was decent. For some reason, even though Google is the king of search, I've never meshed well with it. 

[cweber]

I did try Duck duck Go for a while before just accepting public facing search engines will always gleam something from you. Now I use plugins to minimise that and am careful not to post personal information. Same goes for ChatGPT for that matter.

 

A relative of mine was working on his Bible university PHD proposal and used ChatGPT to brainstorm, capture his though processes and research secular related topics. He said it saved himself literally several months of tedious and laborious trudging through various avenues to accomplish the same thing. The irony was the instructor in charge of the research subject he attended, was citing decades old techniques. It might be that academia still believes individuals have to earn their stripes the hard work way, even if that is inefficient and low-value adding; I mean they had to so why not the students too. I think this is an area where technology has overtaken the institutions.

 

Edited July 22, 2023 by cweber

[Rich]

25 minutes ago, cweber said:

 

I would counter that question with do you expect anything different from Google, who have dominated the search for decades) in this respect?

 

I get how people do not find generative AI helpful. You have to be focused in how you drive the tool. I use it daily for work and other programming interests because it is extremely useful for software development and capturing business requirements and processes. ChatGPT is one AI amongst many now days, but very beneficial. From considerable experience with these AI tools and what I see as an opportunity for improvement in Accordance, I remain convinced AI would be helpful for conversational search and assist with library management. This is especially true for those with a vast investment in library resources. It does not have to be more than that, yet would greatly uplift those two areas (depending on quantity of Texts and Tools) that I find wanting in Accordance.

 

 

I'm with you.  I'm seeing uptake in numerous enterprises.  I imagine almost every Fortunate 500 company is already deeply engaged in discovering how this will be useful in the future.  There is a lot of hyperbole, but this seems to have a bit more traction than a lot of tech crazes of the past 25 years.   

 

I can really see AI in search, synthesis of multiple works, initial research.  It's not magic, but it very much feels like for many categories of software, those who don't embrace AI will be left in the dust (whether their users realize it's AI or not — AI will be powering all the coolest new features).

[Dr. Nathan Parker]

True on search engines. I don't put anything into any "cloud" service that I wouldn't be comfortable the entire world finding out, no matter how "private" and "secure" it claims to be. Any form of hack can happen.

 

ChatGPT was useful in summarizing some works I couldn't get my hands on easily to at least get an understanding of those. I didn't use it for my actual dissertation, but I used it afterward to cross-check some of what I read to see if I understood a high-level overview of some of the material.

 

True on academia. Technological changes are pretty slow in academia. Almost as slow as governments. :-) 

 

I can also see where AI can help with search and synthesis. It's still a little overhyped (I still like sliced bread better), but it'll be useful in numerous applications.

[Mike Atnip]

19 hours ago, cweber said:

 It might be that academia still believes individuals have to earn their stripes the hard work way, even if that is inefficient and low-value adding; I mean they had to so why not the students too. I think this is an area where technology has overtaken the institutions.

 

Yeah, I have white in my beard, so consider that when you read this comment while I rattle on here a little. 🙂

Also, I am not an "academic": my Ph. D. is my post hole digger out in the shed. 🙂

I have to think that perhaps, just perhaps, the idea behind earning a tag behind your name is not that you have managed come up with some good content that wows people. Rather, you have managed to discipline yourself to do what it takes to dig up, sort through, think through, organize, and present an idea. AI can dig up content, sort through it, organize it, and present it (notice that I skipped one! ;-)).

In short, you get a tag on your name for your self-discipline and dedication. If you "cheat" and have a computer do all the "hard work," then you do not deserve a tag. In my mind, "Ph. D." means primarily that "this person is capable of self-discipline, doggedness, and willingness to learn from others," although obviously there are other aspects included. And note that I have the word "capable" in the description. Just because someone is "capable" of doing something does not mean they always use that capability ever after!

Edited July 22, 2023 by Mike Atnip
add two words

[cweber]

My recent discussion on academia may not directly concern Bible study software or biblical texts, but it ties into the broader realm of our educational methods and technologies. This deviation from our usual topics isn't intended to spur political debates, but to reflect on how our learning systems, including Bible studies, are evolving.

 

I aim to contribute to Accordance's continued success, believing that it's essential for any organization to maintain technological leadership and innovate constantly. The recent shifts in Accordance's product releases have urged me to voice my thoughts, and I know many here value Accordance as I do.

 

The rise of technologies such as generative AI, highlighted by OpenAI's ChatGPT, has sparked interest and expanded our understanding of artificial intelligence's potential. This prompts reflection on the state of our educational institutions and how Accordance fits in with that.

 

During my 27-year career from an apprentice to a senior engineer in a leading biotechnology manufacturing industry, I've noticed a shift in academia. It seems to have moved away from fostering the 'type-2 thinking' (the analytical and reflective mode of thought). While I've worked with numerous knowledgeable colleagues over the past 15 years in another industry, it appears such individuals are becoming less common.

 

I agree that we consider how Accordance can keep pace with new technologies without losing sight of traditional academic rigor and intellectual curiosity. It's essential to uphold the principle of 'earning one's stripes' not solely through relentless hard work, but by harnessing efficiencies, and working smarter with the aid of technological advancements.

 

I'm confident that Accordance and its users can embrace tools like generative AI to work more efficiently without compromising intellectual standards. Let's leverage technological advancements to work smarter, not harder.

 

Edited July 23, 2023 by cweber

[Mike Atnip]

49 minutes ago, cweber said:

My recent discussion on academia may not directly concern Bible study software or biblical texts, but it ties into the broader realm of our educational methods and technologies. This deviation from our usual topics isn't intended to spur political debates, but to reflect on how our learning systems, including Bible studies, are evolving.

 

I aim to contribute to Accordance's continued success, believing that it's essential for any organization to maintain technological leadership and innovate constantly. The recent shifts in Accordance's product releases have urged me to voice my thoughts, and I know many here value Accordance as I do.

 

The rise of technologies such as generative AI, highlighted by OpenAI's ChatGPT, has sparked interest and expanded our understanding of artificial intelligence's potential. This prompts reflection on the state of our educational institutions and how Accordance fits in with that.

 

During my 27-year career from an apprentice to a senior engineer in a leading biotechnology manufacturing industry, I've noticed a shift in academia. It seems to have moved away from fostering the 'type-2 thinking' (the analytical and reflective mode of thought). While I've worked with numerous knowledgeable colleagues over the past 15 years in another industry, it appears such individuals are becoming less common.

 

I agree that we consider how Accordance can keep pace with new technologies without losing sight of traditional academic rigor and intellectual curiosity. It's essential to uphold the principle of 'earning one's stripes' not solely through relentless hard work, but by harnessing efficiencies, and working smarter with the aid of technological advancements.

 

I'm confident that Accordance and its users can embrace tools like generative AI to work more efficiently without compromising intellectual standards. Let's leverage technological advancements to work smarter, not harder.

 

👍

[Dr. Nathan Parker]

Good points. I agree that there is a place for AI in various fields, including Bible study. It's interesting to think through the possibilities!

[R. Mansfield]

2 hours ago, Mike Atnip said:

I have to think that perhaps, just perhaps, the idea behind earning a tag behind your name is not that you have managed come up with some good content that wows people. Rather, you have managed to discipline yourself to do what it takes to dig up, sort through, think through, organize, and present an idea. AI can dig up content, sort through it, organize it, and present it (notice that I skipped one! ;-)).

In short, you get a tag on your name for your self-discipline and dedication. If you "cheat" and have a computer do all the "hard work," then you do not deserve a tag. In my mind, "Ph. D." means primarily that "this person is capable of self-discipline, doggedness, and willingness to learn from others," although obviously there are other aspects included. And note that I have the word "capable" in the description. Just because someone is "capable" of doing something does not mean they always use that capability ever after!

 

There have always been ways to find shortcuts, even before the advent of AI, unfortunately. 

 

I personally know a man, whom I will leave nameless. He is now retired and in his 80s. He has a PhD from a Baptist seminary. But he did not write his own dissertation. He did very little of the research for it. He primarily pointed a ghostwriter in a particular direction and let him write it for him. He had an entire career based on that doctorate that he didn’t earn. To this day, he uses “Dr.” and/or “PhD” on all correspondence and anything official—even on his personal checks. And, yet, as I mentioned above, he did not earn that degree. He paid for someone else to do it for him. 

 

All that to say, sometimes letters after a name don’t mean all that much either. There has to be a certain level of character as well. 

 

Shortcuts.

 

Lord have mercy. 

Edited July 23, 2023 by R. Mansfield

[Rich]

2 hours ago, Mike Atnip said:

In short, you get a tag on your name for your self-discipline and dedication. If you "cheat" and have a computer do all the "hard work," then you do not deserve a tag. In my mind, "Ph. D." means primarily that "this person is capable of self-discipline, doggedness, and willingness to learn from others," although obviously there are other aspects included. And note that I have the word "capable" in the description. Just because someone is "capable" of doing something does not mean they always use that capability ever after!

 

I honestly don't disagree with you (and recently wrote an article for an organization on the ethics of AI that spoke a bit to AI in education), but today we don't even blink with people using calculators to do complex math equations, even in school.  (The consequence of which is that many people lack the ability to do basic math in their heads — can anyone make change these days?  ...but that's not the point I'm driving at).  You can now access vast amounts of information directly on the internet without heading down to the library.  We don't think of googling to find a relevant article is cheating.  When does generative AI just become another tool in the toolbox (another calculator)?  I'm not necessarily saying I agree that it is — but I also would say that society relying on calculators and spell and grammar checkers has likely done us a disservice as well, but no one blinks about those today.

[Tim Hall]

19 minutes ago, R. Mansfield said:

I personally know a man, whom I will leave nameless. He is now retired and in his 80s. He has a PhD from a Baptist seminary. But he did not write his own dissertation. He did very little of the research for it. He primarily pointed a ghostwriter in a particular direction and let him write it for him. He had an entire career based on that doctorate that he didn’t earn. To this day, he uses “Dr.” and/or “PhD” on all correspondence and anything official—even on his personal checks. And, yet, as I mentioned above, he did not earn that degree. He paid for someone else to do it for him.

WOW! I am not naive enough to think it doesn't happen, but I can't imagine. And I am working on my dissertation now ...

[ReformedDoc]

@R. MansfieldIf he bought it, then he owns it... so it is his.🤣

[Donald Cobb]

4 hours ago, cweber said:

My recent discussion on academia may not directly concern Bible study software or biblical texts, but it ties into the broader realm of our educational methods and technologies. This deviation from our usual topics isn't intended to spur political debates, but to reflect on how our learning systems, including Bible studies, are evolving.

 

I aim to contribute to Accordance's continued success, believing that it's essential for any organization to maintain technological leadership and innovate constantly. The recent shifts in Accordance's product releases have urged me to voice my thoughts, and I know many here value Accordance as I do.

 

The rise of technologies such as generative AI, highlighted by OpenAI's ChatGPT, has sparked interest and expanded our understanding of artificial intelligence's potential. This prompts reflection on the state of our educational institutions and how Accordance fits in with that.

 

During my 27-year career from an apprentice to a senior engineer in a leading biotechnology manufacturing industry, I've noticed a shift in academia. It seems to have moved away from fostering the 'type-2 thinking' (the analytical and reflective mode of thought). While I've worked with numerous knowledgeable colleagues over the past 15 years in another industry, it appears such individuals are becoming less common.

 

I agree that we consider how Accordance can keep pace with new technologies without losing sight of traditional academic rigor and intellectual curiosity. It's essential to uphold the principle of 'earning one's stripes' not solely through relentless hard work, but by harnessing efficiencies, and working smarter with the aid of technological advancements.

 

I'm confident that Accordance and its users can embrace tools like generative AI to work more efficiently without compromising intellectual standards. Let's leverage technological advancements to work smarter, not harder.

 

Interesting discussion. I've often said that a Ph.D doesn't mean someone is more intelligent than others, just that he's more obstinate. 
 

What was said about the inability to count change, or spell correctly without spell checkers, can also be said about theology and biblical knowledge. Does Bible software help to increase that knowledge or become a substitute for it? I suppose that will depend on the use one makes of it. The better your tools are the more you can leverage the data at your disposal. And for the record, I do think that biblical research has made huge strides over the last few decades, thanks in no small measure to biblical software which has allowed us to research things we probably never would have dreamed of otherwise.

 

How much of that has really helped the Church? Well that's another question altogether...

 

Don Cobb, B.D, M.D, Ph.D 😆😂🤣

[Mike Atnip]

11 hours ago, Rich said:

When does generative AI just become another tool in the toolbox (another calculator)?  I'm not necessarily saying I agree that it is — but I also would say that society relying on calculators and spell and grammar checkers has likely done us a disservice as well, but no one blinks about those today.

I would say this is like Jesus when He sent the disciples out to preach, telling them to not take anything with them. Once they learned that God was all they really needed, then they could take the extra stuff along if they found that it could perhaps be useful here or there. In other words, we can use calculators, AI, and Accordance ... after we learn that ultimately we can fare well without them. I think of the Apostle Paul going on his missionary journeys; he had more success than we often do. And he didn't even have Accordance basic package! 🙂

But the dumbing down of our society is a reality. I could say a lot about this; it pains my heart. I grew up in rural Indiana with parents who refused to put a TV in the house. We had a blast growing up (yeah, literally, we sometimes used dad's well-drilling dynamite in experiments!) While the neighborhood children were watching idiotic (and often immoral) TV, we were thinking for ourselves, building huts in the woods, reloading our own shotgun ammo, learning to repair broken bicycles, trapping raccoons, splitting firewood with a sledge hammer and wedges, and other things that required us to THINK FOR OURSELVES. (When you split firewood by hand, you THINK about the best place to start the wedge because you can save a lot of swings by strategic planning.) It is interesting that the neighborhood boys would get tired of watching TV and come over to see what kind of excitement the Atnip boys were getting into. I've blessed my parents many times in tears for keeping that mind-numbing TV out!

Then after spending about five years in South America and coming back to the US, I sat one day just watching people exiting Walmart. The looks on the faces of the children coming out was sad; so zombie-ish looking, so much of an appearance of a brain that has withered down into just surviving until the next emotionally appealing episode on TV or internet. "God, what's happening to our society?" I cried. Then an older preacher told me that is called "dumbing down of society." [/rant]

I shall return now to my Hebrew studies. And I shall use Accordance. It is a lot easier to mouse over a word than pull my BDB off the shelf and start flipping pages! 🙂

[c. stirling bartholomew]

I have been using Accordance for syntax research for 30 years. I have just recently been augmenting this with AI Chat bot queries on Hebrew syntax questions. I have been formulating somewhat open ended questions by combining keywords for example:

 

Query: Hebrew substantive  construct state linked with  pronominal suffix 

The results are breathtaking. This is going to have serious consequences for ancient language study. What robotics did to Detroit is now upon us in Ancient Language instruction.

[Mike Atnip]

On a whim for another project I am working on, I asked ChatGPT the following IQ question: "“In a lake, there is a patch of lily pads. Every day, the patch doubles in size. If it takes 48 days for the patch to cover the entire lake, how long would it take for the patch to cover half of the lake?”
The response: "Since the patch of lily pads doubles in size every day, we can use the concept of exponential growth to solve this problem.

Let’s assume that on the first day, the patch covers a fraction “x” of the lake. On the second day, it doubles in size and covers 2x of the lake. On the third day, it doubles again and covers 4x of the lake, and so on.

We are given that it takes 48 days for the patch to cover the entire lake, which means on the 48th day, it covers the entire lake (1x). So we can write the equation:

2^48 = 1

Now, we need to find out how long it would take for the patch to cover half of the lake. In this case, the fraction of the lake covered by the patch would be 0.5x.

So, we need to solve the following equation for “n”:

2^n = 0.5

Let’s find the value of “n”:

n = log2(0.5)

n ≈ -1

Since “n” represents the number of days it takes to cover half of the lake, and we can’t have a negative number of days, we round the value up to the nearest whole number:

n = 1

Therefore, it would take one day for the patch to cover half of the lake."
Flunked! 😂
Perhaps one of those warnings should be put on ChatGPT: "Use only under adult supervision."