E41: Can I Use AI, or Will AI Use Me? - podcast episode cover

E41: Can I Use AI, or Will AI Use Me?

Apr 25, 202326 minSeason 4Ep. 41
--:--
--:--
Download Metacast podcast app
Listen to this episode in Metacast mobile app
Don't just listen to podcasts. Learn from them with transcripts, summaries, and chapters for every episode. Skim, search, and bookmark insights. Learn more

Episode description

AI is everywhere now. 

You might now be fully aware of it, but you carry it in your pockets—our phones are built to understand users through data usage and the apps that we download make use of AI technology. In the workplace, the competition shifts not only in making sure employees are equipped, efficient, and accurate but that they are also fast and productive—and that’s where AI takes the center stage. Imagine: A report which takes an employee to write in one to two hours can be generated by AI in less than 5 minutes.

Truly, what AI can do is astounding. But it’s also scary. Comes with the rise of AI is a question we don’t want our lips to speak of but is an impending idea that must be addressed: are humans going to be replaced by robots or AI?

Learn how AI and humans can co-exist in this episode as I discuss:

  • The role of artificial intelligence in a service-based industry
  • The dangers of an exposed IP to AI tools like ChatGPT
  • How to protect intellectual property from potential AI and other human infringers

I received many questions regarding the concerns about the rise of AI robots in the workflows. Feel free to reach out if there are anything you’d like us to look at to help you protect your assets and intellectual property.

Connect with Erin and find the resources mentioned in this episode at hourlytoexit.com/podcast.

Erin's LinkedIn Page: https://www.linkedin.com/in/erinaustin/

Think Beyond IP YouTube Page: https://www.youtube.com/channel/UCVztXnDYnZ83oIb-EGX9IGA/videos

Music credit: Yes She Can by Tiny Music

A Team Dklutr production

Transcript

Hello, ladies. Welcome to the Hourly to Exit podcast. today we tackle artificial intelligence. Now, there's no one more surprised than me that artificial intelligence has become part of my life. I am a proud technology Luddite. if it was up to me, I'd still be using a Blackberry. To me, they're, the. Things ever. but there's no denying the fact that AI is everywhere and we can no longer put our heads in the sand regarding this. there are some dire concerns regarding it.

I know that there are some tech leaders who have gone so far as a call for a pause in AI development. I did not know. I'd agree with Elon Musk. Pretty much anything. But I do agree on this point, today, in fact, as I record this, there was an article about an episode of the Joe Rogan experience that was completely AI generated, and apparently I've not listened to it, but according to the article, you could not tell that it was.

The true thing, the true, recording, live recording, and that's kind of scary. I have seen online. completely. AI created YouTube videos, so they use, image and a voice stamp, and they're able to create an entire video. I will say that the AI generated YouTube video that, I saw looked like it had a few problems, but it's just a matter of time before it will be truly indistinguishable from live recordings. And so there are concerns. Now I am not going to address those issues.

They're beyond my expertise. What we're gonna talk about today is the issue that people have been asking me, which is, what impact will AI have on my intellectual property? people be, will I be replaced by? So the answer to will you be replaced by ai? The answer to that is no. as an expertise based service provider, I understand your concerns about the impact of AI on your intellectual property, but they aren't.

Any greater than kind of the general of accessibility of our expertise online these days. So let's, just start from the beginning. So let's look at how AI works. when an ai, System like chat, g p T. Now I'm going to talk about chat, g p t mostly today. That's the AI that I'm familiar with. It's the one that is publicly available. probably the most advanced publicly available one.

There are others like Bard I believe, and there's a couple, that aren't as advanced as chat g b. So if you've been fooling around with it, the way I've been fooling around with it, you've probably been doing it on chat, G B T and that is, it generates responses from available data. So I understood that there was a source of data. That they copy from the internet and then it's saved by, open ai, that's the creator of Chat, G B T. and that the date of that saving is September, 2021.

So it doesn't know about anything that's happened since September, 2021. And so to understand this a little better, I asked Chad, G b t I don't know if. Had these conversations with them. I encourage you to go on if you haven't and just ask some questions so you can see how it works. It's really. interesting. It is different than a Google inquiry. It absolutely seems to be talking to you that we call them conversations or chats.

so I asked chat g p t for confirmation about, what's the last copy of the internet that it has, and so it's response. As an AI language model, I do not browse or access the internet in the same way that humans do, and I do not have a quote. Unquote copy of the internet. Instead, I rely on my preexisting knowledge base, which was last updated based on data and information available up to September, 2021. So, this was, as of yesterday, which would be, April 13th, 2023. And so that will.

in the near future understand that the next generation of AI will be able to collect data in real time. So it won't just be the preexisting knowledge base that was loaded into it. It will be able to comb the, internet live to find its answers. So obviously that will make it much more powerful. So the implications for you as an expert. First AI will be using your ip. AI will be accessing any information available on the web, so currently available.

Assuming that it was downloaded, as of September, 2021. So the interesting thing about the answer that it gave me that it has a preexisting knowledge base. Well, that means someone had to load that on there. It had to decide what information they wanted to input. So maybe your IP got into its knowledge base, but maybe it didn't. But when AI is able to just kind of comb the, internet live, then obviously it will be able to access whatever, is publicly available on the.

And so that would include, publications on your website and other platforms. And so while that may sound a little bit terrifying, that happens without AI right now. I mean, Google and other, search engines does live comb the internet and can access your intellectual property right now. So it can find. what you put on LinkedIn, it can find out what's on your website. It can find out, copies of, your speeches on YouTube so people can do that right now without ai.

And so if someone live, without any intermediary, goes on the internet and sees that you wrote something and decides to write something similar, We hear all the time about people are copying my ideas. There's nothing new about that. so AI will be able to see what you're doing and incorporate it into an answer that they provide But what we need to understand is that in both cases, being inspired to write something similar or even stealing your ideas is not copyright infringement.

copyright infringement is the same, whether it's a human reading your stuff and taking it, or AI reading your stuff and taking it Copyright infringement means that they have to have ex to violate your exclusive rights as a copyright. So as a copyright owner, these are your rights. You have the exclusive right to reproduce, distribute, publicly, perform, and publicly display your works.

So anyone who engages in any of these activities without your permission, they would be liable for copyright infringement. So copyright infringement means that as the copyright owner, you can show that one, you are the owner of a valid copyright in the work in question two, that the infringer has actually copied your work or a substantial portion of it. And three, that the copying isn't protected by an applicable defense such as fair use.

if you have a question about fair use, I have, written about it and done a LinkedIn live about it, which you can find. but copyright infringement again, is not about stealing your ideas, but it is about actual copying. And so if you can prove that they actually copied your work, then they would be able to take, you be entitled to damage. And I'm not gonna say that proven copyright infringement cases is easy. It's not, it's complicated and it typically involves a fair amount of legal fees.

So you do need to make sure that it is valuable enough to you to devote those types of resources to. Prosecuting an infringement case. So, but if you are concerned about the use of your IP by a third party, whether that third party is human or an AI tool, then you need to take the steps required to protect it. So if we go back to proving infringement, you have to be the valid, owner of a valid copyright question. Chain of title regarding ownership starts from the time of creation.

So you'll hear me talk over and over again about the importance of your contracts and making sure that you own things. If you didn't create it personally, if the third party created it for you, if you created it in connection with an engagement, then ownership will depend on what those contracts. And then, when you are the sole owner, then registration. So these are the two ways we make sure that we have a defense.

we can, enforce our copyright against third party infringers ownership and registration. and then, again, it would have to be a copying of your work and not. Taking the idea of your work. So it is not in your power to stop someone from accessing your IP on the internet by humans or by ai, but it is in your power to put the pieces in place to put yourself in the best position to enforce your rights if it comes to that.

So on the other end of that is when you're using ai, And AI is doing work for you. It will be accessing other people's intellectual property on the internet when it's creating its responses. Now, when you, go ahead and, play with it, you will see it will send you answers. It does not provide any attribution. So that is different than if you do a Google search where you know exactly where the information is coming. Here. It just goes wherever it goes and it's.

Knowledge base and finds answers, and then it, sends it back to you in a, if you asked for to write an article for you or to write a script for you, it will present it in that form and, typically without attribution, so you won't have any way of knowing where that information is coming from. So, That puts you in danger of being a copyright infringer if you don't know where that came from.

So I asked chat, G B T what to do if I'm worried that it's response infringes a third party's copyright, and its response was this. As an AI language model, I provide responses based on my preexisting knowledge and language capabilities. However, I am not able to review or assess specific content or material for potential copyright infringement.

If you are concerned about potential copyright infringement related to your own content and material, you may wanna consult with a legal professional who can provide guidance and advice specific to your situation. They can review your material and assess whether any potential infringement has occurred or is likely to occur. In general, it is important to be aware of copyright laws and to obtain permission or authorization before using or reproducing any copyrighted material.

Additionally, it is advisable to seek legal advice if you are uncertain about whether your use of someone else's copyrighted material may infringe on their rights. Well, in other words, user beware. I mean, that's all very fine to say, but if you have receiv. A response that doesn't provide any type of attribution. All we know is it came from chat GPTs preexisting knowledge base. Like how does that tell me what the source is?

How do I know if I need to, if this is a potential copyright infringement? Again, you know, copyright infringement requires copying and not just inspiration or. taking an idea. So did they completely take a whole chunk of somebody else's, work and stick it in into your response? Or did they, take a variety of sources and kind of put them together in a way that creates something new?

So if chat G p t Plagiarizes from source materials, You won't know that unless you do your own due diligence and you are the one who'll be liable for copyright infringement, not chat. G P T. So if you plan to use AI for your work, you need to use some sort of plagiarism check, like Grammarly, and you need to track your process carefully to demonstrate your unique input when you're creating content. So you can't just, copy and paste what comes out of your AI chat.

There also are some privacy issues, so. When I asked chat G b T, whether or not the information I provide to it becomes public. The response was, your information is not made public and is kept confidential. However, if you go to the fax page, it warns you don't share any sensitive information in your chats. Your conversations may be reviewed by RA ai trainers to improve our.

So remember the way that AI works is that what you put into the chat becomes part of its knowledge base, cuz that's the only way that it can continue to improve your responses. So again, if you go in there and, maybe you put in. A draft of something and ask it to improve the draft, it will, maybe you put in notes, put some bullet points in, and ask it to create an article based on these bullet points. It will use your bullet points.

It will use its preexisting knowledge, and it will come up with, a. Something in article form and then you say, eh, and you ask it to regenerate and then it will go back and it will improve on it. And you'll see that each time you ask it to regenerate, it gets better. Cuz it kind of goes deeper into its, knowledge base and it's got, what you asked it to use as a starting point and it gets better. So somehow, What you have input is being added to its knowledge base, so it can give you better.

So, as a lawyer, you know, my ethical duties means that I cannot share any confidential client information with a public, AI tool such as chat, G p T. So I recommend the same diligence for you with respect to your confidential information, and of course, your client's confidential information. You need to protect it and assume that at some point, if not now, then certainly later that anything that you put into a chat will be accessible.

So finally, there is kind of the issue about ownership of content that is generated by ai. So the US copyright office is working out. Whether content generated by AI can be protected. chat, G B t, says that the answers are owned by the requester. but that. Make some assumptions that I don't think you can rely on. Again, we don't know the source of the answers of the responses. So without knowing that, how can you know that you own it? copyright law does not protect ideas.

It doesn't protect facts, it doesn't protect information. so a lot of AI generated responses fall into those categories. However, there are. creative uses for ai. sometimes, you can ask it to write a poem or you can ask it to write a script and you can, ask it to write songs. And certainly there is AI that can create, images and paintings that would be considered original and creative. So there are some real valid questions about ownership of AI generated content.

I describe intellectual property as the product of human intellect, and AI is not human, at least not yet. At the beginning of ai, there had to be humans. There always has to be a human at the very beginning of any idea, right? at least a recorded idea. so this is just, another one of these conundrums that AI presents to us that will need to be worked out as AI becomes more prevalent.

So, When we talk about your expertise, your intellectual property and the dangers of it being exposed in, more scenarios through the use of ai. I talk about the distinction between a cook and a chef. a cook can follow a recipe and produce edible and maybe even delicious food. However, a chef has that skill, the experience, the creativity, and the talent that makes people wanna pay the big bucks to go join, to pay them to, have that experience at their restaurant.

So this is the same with your expertise, so an AI solution? Yes. It can replicate, you know, run of the mill. Of what you do, it can scrape the internet for freely available content. AI is basically a generalist though, and it doesn't understand the nuances the way that an expert does. And when it's scraping the internet, it won't be able to differentiate between, what is thought leadership versus what is just someone's random ramblings, on the internet. And so that is why today.

You'll get some nonsense answers from AI cuz it, can find something that matches your keywords, but it can't tell whether or not this is well-researched information or if it's just something, someone's musings so it cannot, kind of synthesize. Experience the way that you can, I mean, experience and being able to kind of pull on that is a uniquely human skill. So it can only pull on what has been recorded and it can't always apply. Context either.

So as a chef, as the expert, you know when to go with a standard ingredients or when to add something else, you can taste it and go, eh, it's too sweet, it's too salty, and know how to tweak it to make it taste better. Another metaphor is with respect to, the doctor-patient relationship. Yes, AI can ask diagnostic questions, it can review lab results, and it can provide a, diagnosis based on that and propose solutions based on that. What it can't do is it can't observe the patient's demeanor.

is sweating, is their leg shaking? can they notice that maybe it's, someone who has a parent in the room with them and asked, you know, can I talk to, your son alone? they can't ask the follow up questions. They can't put him at ease. Sometimes having your blood pressure taken and you're like, you know, why don't you lay down for a second and we'll do it again. These are a number of things that require kind of the human touch.

And some people will accuse doctors of not having a bedside manner, but most doctors do have a bedside manner. AI can never have a bedside manner, right? And so those things where you as an expert can use your instinct. To take, what you know about a client, what you know about their, problems they had with their last consultant. these are things that AI cannot do.

So ai, yes, it can be an excellent support for your work, but only you kind of know how to tweak things and make those little shifts that come from years of experie. So I had one more question for chat and G B T and I asked it, can AI take the place of a management consultant? And here was its answer. AI can certainly provide valuable insights and analysis that can be useful for management decision making, but it is unlikely to completely replace the role of a management consultant.

Management consultants bring a range of skills and expertise to the work, including strategic planning, business analysis process, improvement, and change manage. They often work closely with clients to understand their unique challenges and provide customized solutions that meet their specific need. Management consultants also typically have strong interpersonal and communication skills, which allow them to work effectively with a wide range of stakeholders.

While AI can provide powerful data analysis and insights, it may not be able to replicate the nuanced and contextual understanding that a skilled management consultant can provide. Additionally, AI may have limitations in its ability to understand and respond to complex human factors such as organizational culture, individual personalities, and political dynamics. Oh, that is so.

That being said, AI can certainly be a valuable tool for management consultants helping to automate certain tasks and provide additional insights that can inform the recommendations. Ultimately, the most effective approach is likely to be a combination of AI and human expertise working together to provide the most comprehensive and effective solutions for. Yeah, that is absolutely true. So I am excited by AI's possibilities to support us as experts in our work. It is absolutely a time saver.

It is a wonderful resource, but ultimately, without human touch, AI has its limitations, so make it a fantastic tool. As part of your toolbox, how you provide value to your clients, how you make your business more efficient and therefore more profitable. But I don't want you to worry about it replacing you as an expert. The key to working with A I protect your original ip, protect it through documentation.

Through registration and when you're using AI as a resource, make sure you are doing your diligence to make sure that it has not plagiarized somebody else's work. So if you have any questions or concerns about how AI impacts your ip, you can contact me for consult. And also don't forget the IP tracker as a way to determine what steps you need to take to protect yourself and your ip.

Transcript source: Provided by creator in RSS feed: download file
For the best experience, listen in Metacast app for iOS or Android