DeepSeek Integration
Hello, it would help to have search AI answers using deepseek.
Currently the answers are too short.
Also, it would help to use deepseek, given that this has bigger context window.
Comments
-
0
-
Didn't Deepseek just get caught stealing data?
You want to be getting your Christian instruction from thieves?Or you want our Christian knowledge to be open to secular molestation?
Or a direct wiretap to the Chinese government? (The USA may be banning TikTok for a good reason, just as they banned Kaspersky)0 -
Thank you for your comment! I agree with many things you say, but I think there is some misunderstanding.
DeepSeek is open source so the searches in Logos are not accessible to the Chinese (unless their API is used, which is not what I am advocating for).
Whether Logos uses it or not, it does not affect China at all (if self-hosted). By using DeepSeek, it does not financially contribute to China because it is open-source too (aka free). I am not calling for Logos to use DeepSeek nor DeepSeek's API. If Logos self hosts DeepSeek, it all becomes very private.
The integration does not have to be with DeepSeek too. Integration with open source Llama would be also helpful because of their greater context window. It would only help Christians.
Also, the AI Logos is currently using comes from a secular source.
0 -
The reason I wrote what I did is because I've seen this before.
On another forum the execs of AI firms are sending people to try and persuade the devs to integrate AI systems in their software. Giving access to peoples' personal files, cause that AI is so very hungry.I couldn't find what AI Logos uses. But the principle is that its data source is sandboxed to FaithLife's considerable literature catalog. So it's not possible to provide information that we don't already agree with. And I sure hope it's not possible for complete strangers with undisclosed worldviews and motivations to know everything that I study.
…
Microsoft is trying to convince the unsuspecting masses to allow their client side scanner (that exists to defeat end to end encryption) onto their computers. Using soft colours, softly spoken talking bobble heads and cute animals to make it seem as safe and friendly as possible; while under that white washed veneer is the most horrifying Orwellian capture and control mechanism that's ever been seen on this Earth. We used to call this spyware, and now it's called AI. It's a weapon of war.
Nice try, Microsoft. I'm not going to die if I have to struggle out a name for my cat. Or perform extremely basic tasks that men have been doing just fine since there has been men. Thanks anyway.
Those who control the superior AI system will rule the world. That's what this race is about. That's what it's always been about. Money, sex, power. Period dot. It has nothing to do with curing cancer, or solving "the mysteries of the universe" that don't even exist anyway.
"I was born at night but it wasn't last night."
- Dr. Kent Hovind0 -
I couldn't find what AI Logos uses.
that is because they are carefully "generic" making it easy to change providers in this rapidly changing environment.
Orthodox Bishop Alfeyev: "To be a theologian means to have experience of a personal encounter with God through prayer and worship."; Orthodox proverb: "We know where the Church is, we do not know where it is not."
2 -
That's smart for them to do that. Not so good for us.
Every company out there has a security and privacy statement of = trust me bro.
We won't take your data and sell it cause = trust me bro.
We won't get hacked and lose your data to the dark web cause = trust me bro.I will give FaithLife the benefit of the doubt, but not any random AI company they might partner with.
Don't forget anyone on the internet are complete strangers, some are even bots, and there is no such thing as a cloud, only other peoples' computers.0 -
I don't believe that Logos is sending any personal data. The only reason they would have to send any of our documents would be our sermon/homily texts. Do you have any reason to believe otherwise?
Orthodox Bishop Alfeyev: "To be a theologian means to have experience of a personal encounter with God through prayer and worship."; Orthodox proverb: "We know where the Church is, we do not know where it is not."
0 -
It doesn't matter what the FaithLife app does. If they integrate another technology, then the behavior of that external technology is out of their hands. They can require certain behavior with legal contracts, but even then we have many examples of tech companies completely disregarding the law. Especially with AI.
Currently when we use the AI search it sends the question to their side and it gets processed there (as far as I understand it). So no AI system exists locally. It can't possibly take anything if it isn't even on the computer. But not drawing a line in the sand is a slippery slope. Things can change in the future, as it always does. This is the doctrine of increments. "Boiling the frog slowly". Taking a little bit more data, one step at a time. So slow that nobody notices. But again I mentioned that I give FaithLife the benefit of the doubt. And no other tech company out there can be trusted. They've proven that fact again and again.
0 -
Having worked in statistics and enterprise systems IT from 1968 until retirement, having a brother (hardware), nephew (military grade security), son-in-law (Microsoft level networking) and a grandson (AI security) in the field as well as having an obsessive interest in computational logic, there are many holes in my knowledge. But I do know that Logos takes advantage of portals that allow you to feed text to the AI engine as is evidenced in the synopsis drawing only from your Logos library. AI cannot take nibbles of data from Logos, however it can, and does, take a stream of data that Logos sends.
Orthodox Bishop Alfeyev: "To be a theologian means to have experience of a personal encounter with God through prayer and worship."; Orthodox proverb: "We know where the Church is, we do not know where it is not."
1 -
I do not think using Deepseek or any other AI program would be beneficial for all of us using Logos. In my opinion, Logos should create their own AI system that is only compatiable to Logos. The people using this system should be only the subscribers that are using Logos.
0 -
For Logos to create it's own AI it would cost millions. I would rather like to see that money in ressources.
I believe AI can be very beneficial for simple both for the (simple) Bible Reader and for the theology professor.1 -
All AI systems are stealing data big time from the Internet to train their systems.
2 -
Well, now, you have your 'stealing'. And then, you have your 'STEALING' (depends on the stealee).
"If myth is ideology in narrative form, then scholarship is myth with footnotes." B. Lincolm 1999.
0 -
Ya. The problem is you wouldn't be able to have AI at all without feeding it huge amounts of data, at first at least. After initial training it can remember what is the correct course of action, so it doesn't need to calculate that again.
I'm not sure how many people would allow their data to be used, so it's possible the technology wouldn't exist at all if it hasn't been stealing it this whole time.We just really need a copyright clamp down on these generative AI parasites. It needs to stop.
At the very least a requirement of citation for AI content.1 -
@ASUNDER I agree about the citation thing. Something mandatory that lets everyone know it's AI generated.
On the other hand I see it like with any new technology. At the beginning there are "birth pangs" and "childhood diseases". Some have more, some have less. And with every new thing people need time to really get to know it, start to basically understand it's function, have experience, it needs testing, etc.
But that does not make watchfulness and a critical mind irrelevant!1 -
Asunder, you're hopeful. My observation, is that humans choose convenience. Get it and go. Sitting around checking it, right.
"If myth is ideology in narrative form, then scholarship is myth with footnotes." B. Lincolm 1999.
2 -
"With every new thing" doesn't apply here. This isn't the same as previous technologies.
In the past they invented machines to improve physical output. Now it's mental output.
The machines didn't steal from anyone. It was given raw materials that the owner already purchased legally.
Now the raw material has been stolen from everyone. And this same "AI" is going to have the power of filtering people based on their social credit score. It's already in legislation to protect children from pornography, but that's not what it's for. The same data that is used to protect is also used to control. This is not the same as other tech.Actually I'm cranky about this because I'm not hopeful.
You're right about people and their convenience. Canva is the most godawful horrible thing I've ever had the displeasure of using. And it's by far the most popular image editing program. To the point it's now gobbling up other graphic art companies.
"I will take this". (notice they produce content).
This is the problem. Not that AI brainrot is pure cringe, but that the market doesn't care. So the cringe gets rewarded and incentivized. There are now 600k RedBubble accounts. To the point people now say it's a waste of time to even try. If you spent a hundred hours drawing a work of art you'll get zero sales cause that edgy teenager stole some trendy thing with AI and takes $5k a month for an hour's work.They really need to make selling AI content for money illegal. This is a necessary thing to save our internet.
When we have an infection we are told to take the medicine right away, or else it will grow and kill us. This AI is a cancer on the world.0 -
It's not the AI (OpenAI, Groq, Deepseek, Gemini) that is important. ANY of those can provide the computational framework. It is the LLM (large language model) that the AI pulls from that is critical.
For example - if you go to a company's website and there is a chatbot in the bottom right. That chatbot might be powered by an API integration with OpenAI. But the only data it accesses is what is on the website and/or what the company has put into the Large Language Model for the AI to pull from.
That is what FaithLife is doing. That way, the AI in Logos only pulls from valid credible sources already vetted for authenticity.
If we want more than that, just have ChatGPT, Claude, or your AI model of choice open in another window and ask it there.0
