Ep 79: The Importance of Ethics in AI

 

Episode Categories:

 

Video Insights

Overview

The advent of AI has revolutionized numerous industries, promising unprecedented opportunities for growth and efficiency. However, as AI continues to gain prominence, the importance of ethical considerations cannot be ignored. Ensuring responsible and transparent use of AI is crucial to protect individuals and businesses from potential harm. In this article, we will delve into the significance of ethics in AI and how it directly influences business success.


The Need for Ethical AI:

In recent times, AI has become an integral part of our lives, impacting various spheres such as healthcare, finance, and transportation. The ethical implications associated with AI development and deployment are vast. Business owners and decision-makers must prioritize incorporating ethical considerations into their AI strategies to foster trust among stakeholders and protect against reputational risks.


Building Trust and Transparency:

Transparency is fundamental to establish trust in AI systems. Organizations that embrace transparency by clearly communicating how AI operates and handles user data are more likely to win the trust of customers, clients, and regulators. Implementing stringent policies that prioritize data privacy and security inspires confidence among users and demonstrates a commitment to responsible AI practices.


The IBM Approach:

Companies like IBM have recognized the significance of ethical AI and developed robust frameworks to govern its use. By adopting an "AI by Design" approach, IBM ensures that ethics and responsible practices are integrated at every stage of AI development. With pillars of trust, transparency, privacy, fairness, explainability, and robustness, IBM sets an industry standard for ethical AI implementation.


Addressing Disparities:

AI has the potential to amplify existing inequalities if not deployed responsibly. Bias in data and algorithms can disadvantage certain groups, leading to unfair outcomes. Business leaders must proactively address these disparities by analyzing and mitigating bias in the data used to train AI models. By actively promoting fairness and inclusiveness, businesses can ensure AI serves as a tool for societal progress rather than exacerbating existing divisions.


Ensuring Accessibility:

While the accessibility of AI has improved, some challenges remain. Affordability and availability of AI technologies continue to be barriers for smaller businesses and individuals. As AI becomes more accessible, business owners and decision-makers must strive to create a level playing field by providing opportunities for all to harness the potential of AI. This inclusiveness can drive innovation and foster a diverse and thriving business environment.


Conclusion:

As AI continues to shape the world of business, organizations must navigate the ethical landscape with care. By prioritizing transparency, fairness, and an inclusive approach, businesses can build trust among stakeholders, mitigate risks, and unlock the full potential of AI. Embracing ethical AI practices is not only morally imperative but also vital for long-term success and sustainability in the evolving era of technological innovation.


Topics Covered

- Concern about companies not prioritizing ethics in AI
- AI's potential in education and its real-world applications
- Ensuring individual responsibility and protecting personal data in AI usage
- Dismissing extreme fears of AI leading to mass destruction or job loss
- Accessibility and awareness of potential dangers and limitations of AI
- Examples of misuse of AI as learning opportunities
- Importance of using AI in an effective and optimized manner
- IBM's history and hopes for their approach to AI
- Concern about AI disadvantaging certain groups of people
- Ethical concerns regarding programmatic ads generated by AI
- Existence of bad actors in the market seeking to monetize AI
- Surprising lack of awareness about the advancements and impact of AI
- IBM's commitment to ethics and responsibility in AI
- AI ethics board and privacy board at IBM
- AI by design process and pillars of trust, transparency, privacy, fairness, explainability, and robustness
- IBM's stance against military applications and surveillance technology
- AI's role in HR tools and resume screening
- Partnerships with Vatican and University of Notre Dame for AI ethics
- IBM's support for precision regulations and guardrails for AI technology
- Watson X platform for enterprise AI, including Watson X.ai, Watson X.data, and Watson X.governance subsets
- Importance of accessibility and leveling the playing field in generative AI
- Updates on AI-related news, such as The New York Times' terms of service and funding for Anthropic
- Amazon's use of AI summaries in product reviews
- Excitement about the accessibility and benefits of large language models


Podcast Transcript


Jordan Wilson [00:00:17]:

How should we be using generative AI? It seems like there's no shortage of of tools and softwares that can help us do our jobs better, but we don't really ever talk about the ethics behind it. So that's what we're gonna be doing today on everyday AI. So welcome. This is your daily live stream podcast and free daily newsletter helping everyday people keep up with what's going on in the world of generative AI. So Yeah. We can always read about it. but this is where we actually talk about it. And we say, here's what it means to to your business, to your company, to your career. So extremely excited today. we have, a senior leader from IBM joining us.

Daily AI news

But before we get into that, Let's talk real quick about what's going on in the world of AI news. Let's make sense of this. Alright. So first story, The New York Times just says just said that companies can't use its site to train AI models. So they updated their terms of service kind of barring, different AI chats from crawling their site and using that information, or, sorry, restricting end users to use that information. So, very interesting timing, especially as OpenAI kind of, released so to speak, their GPT bot, which told companies how they crawl their website and use their information. So this could talk about this one for hours, but keep your eye on that because I think it is pretty, pretty noteworthy.

Alright. Next piece of news, ChatGPT competitor raked in a huge fundraising round. So anthropic, which makes the, LLM cloud 2 recently just announced a $100,000,000 investment from a Korean telco company, SK Telecom. So if you don't know, anthropic was founded by former OpenAI Leaders, the makers of ChatGPT, And they recently raised about 3 months ago, $450,000,000. So a lot of people haven't even used anthropic in their cloud too it's definitely worth checking out if you haven't already and, you know, pretty big fundraising news here, bringing in another $100,000,000 investment.

Alright. So last but not least, in our AI news today, Amazon just just noted, just less than an hour ago, just released that they are going to be using AI summaries, in reviews. So they actually just started to roll this out across the US to select users, but more or less here's what it means. You are going to see AI summaries of product reviews, which I think is fascinating because so many times, at least for me, you know, you know, whether it's on Amazon or any other sites, I spend so much time reading reviews So we'll we'll see how accurate these, AI summaries are, and there will also be a product insights feature that kind of highlights common trends found in reviews. So, maybe I'll spend less time reading reviews and more time spending even more money on Amazon. 

About Moe Alo and IBM

So I'm excited. I'm excited to talk AI in ethics. It's it's something we don't talk about a lot. Don't worry. It's not just gonna be me blabbing. it's not just gonna be me blabbing today. Have a great guess. So Let's bring in our guest for today, Moe Alo, who is the, senior sales leader, a senior sales leader at IBM. Moe, thank you for joining us.

Moe Alo [00:03:46]:

Jordan, I appreciate you having me. I'm excited to talk to you. I'm a big fan of your work.

Jordan Wilson [00:03:51]:

Thank you. Hey. Appreciate that. So As a reminder, for those of you joining us live, like Nelson here who's saying greetings, this is a live show. This is unscripted. So if you have a question for Moe, just about AI and ethics or if you're just interested even in, hey, what's IBM doing in this space? Make sure to drop a comment like Michael who says cloud rocks. I like cloud as well. so so, Mo, real quick, just tell everyone a little bit you know, what you're doing, at IBM, what's kind of your role as a senior sales leader at IBM entail?

IBM's importance to AI


Moe Alo [00:04:23]:

Sure. in at IBM, my role is I cover financial services. So I, work with a set group of accounts. and, you know, we we believe in consultative process and the collaborative approach. And so what we do is try to learn as much about our clients and their pain their pain points and then, you know, work with them to see how we can assist them. as far as AI and you probably are aware of this. IBM has been, you know, at the forefront of AI for a long time. I'm something that we're proud of, and and we are as excited as every else is with all the new advancements that have been going on and how we can, satisfy the needs of our clients and make sure that they're keeping up and also in in appropriate ways?

Jordan Wilson [00:05:14]:

Yeah. It's it's something I don't think we even talk about enough on the show. is is just IBM's role in this because I think if not for IBM's, initial investment decades ago, in OpenAI. you know, Watson, Deep Blue, all of these, you know, now that we kind of consider quote unquote old school, advancements in AI. I don't know if we'd be here. Like, what do you think about that, Moe, and and how, how important do you think that IBM has been to even where we're at now, even with other companies?

Moe Alo [00:05:47]:

Well, I I think you're right. You're spot on. you know, IBM has invested into the research and into the process of developing a an AI system and an AI that works and is trusted. you know, deep blue and, Watson, you know, going back to, you know, the, was it Gary Casperov, the the chess player that was defeated by a deep blue and then Watson, 1 jeopardy. And so it's kinda like the moon landing. Right? Those are kinda like the first markers in in in making, you know, a large publicity for AI and the and what large language models can do and what what that does is gives us, you know, a basically 2 decades of of working with AI models and fine tuning them and,

Jordan Wilson [00:06:43]:

learning how to best deploy them So, yeah, that's something we're really proud of. Yeah. Yeah. And timely timely comment, from from Miguel joining the live stream here saying growing up when I saw Watson doing his thing in jeopardy, I was like, I want this for my personal life. And now we we we we have this for our personal life. So so Mo has someone that's been involved in, you know, computer and and technologies for a long time. Can you just speak to now, like, you know, even Mikael's comment here, we have this in our personal life.

Moe Alo [00:07:13]:

Yeah. I mean, it's funny. Right? we still don't maybe a lot of us don't realize it and don't see it our way that way, but it is. And and Jordan, I saw your, pod recently on, you know, how you can personalize ChatGPT, right, to learn how to write like you, which I think is awesome. I mean, I think that's incredible that we can actually do that and the way that, you explained it made it very easy for people to use. Look, AIs is advancing at such an incredible rate and such a phenomenal rate everybody, is being able to use it and learn from it. and Jordan, you do a daily show, right, 5 days a week. You take sun Saturdays and Sundays off. Thank god. But you could probably do an hourly show. I mean, all you have to do is be on Twitter, be on LinkedIn, and and follow, you know, people that are that are, you know, pushing AI and and just, relating the latest developments. And, you know, there's there's so much, that what you touched on earlier ethics is obviously a a big, concern, not not just for IBM, but for all the the large, companies out there that are involved in AI and developing

Jordan Wilson [00:08:27]:

Yeah. And kind of finally bringing us to our topic here, you know, ethics. Because one thing that I find interesting is a lot of times bigger con bigger companies, so enterprise level companies that IBM would normally be working with are most set up or most, or maybe set up the best to leverage generative AI. But a lot of times, the the the larger enterprise companies at times maybe the the last to to adapt. So what's what's just your overall take on ethics in generative AI. How should we be using it? Should, you know, whether you're working at a, you know, a large company or small company, should we be using generative AI in all aspects of our work? Like, what are your what are your thoughts on that? Because, you know, it's it's a deep conversation and people have opinions, but you you know, what are your thoughts just, you know, given your background, though.

How should we be using AI?


Moe Alo [00:09:20]:

Yeah. so let's, let's try to unpack it and categorize it. Right? Should we, let's say, should we be using AI? Absolutely. I saw your your your pod regarding should change your BTD band in schools. And I I love the approach that you took. Why would we, give our students, a disadvantage when they're gonna be able to use it in the real world, learn how to use it. You know, what's the biggest what's the the most common statement that that I've heard at least is, you know, John humans are not gonna be replaced by AI. They're gonna be replaced by people that use AI. Right? But the reason that thing that that that statement took off is because it is true and it makes sense Right? we have to learn how to use it. You know, it's like, I think you had mentioned, like, banning the internet. Right? etcetera. And so it is important for everyone to have an understanding of it. not only for them to learn, but also so that they can know the trouble spot They can go through trial and error. you know, there's there's a lot of stow let's go back to why it's important ethics in AI. I think that there's a personal responsibility, right, user ethics. how are we as individuals using it? there's the responsibility, right, protecting, data of the enterprise of, personal data that that enterprises collect. the outcomes of AI. Right? Are we gonna be, you know, creating, you know, the the big fear that everybody has, you know, the the Terminator movie type AI everybody dies or, you know, people lose their jobs. Obviously, you know, that's not that's not the case. That's a that's an incredible stretch. And then accessibility, I think is important. You know, everybody needs to have access. Everybody needs to be able to benefit from the the tools that are out there. and I think, as as users, we are developing scar tissue and muscle memory when it comes to, you know, the the dangers of AI. Just kinda like Photoshop when that first became popular. you know, we kind of figured it out. It didn't take long to wear, okay, that looks photoshopped, etcetera. The same goes for deep fakes. I know that there's a lot of talk about watermarking, AI generated content and there was a recent executive order or or a a agreement on 8 principles, of AI that I'm sure you've covered. So I don't need to tell your your audience has to be the educated. They there's like nothing new I can tell your audience. I saw a good job with the job you do. hallucinations. We all know about it. We've heard of we've heard of, you know, situations where people, you know, got caught up or, you know, lawyers using AI that made up cases and and know, those are things that are unfortunate, but I think that's part of our process as everybody to learn, you know, what exactly, AI can do, but how should we be using it? I think we should be using it in the most, effective, optimized way and doing our part to have guardrails, you know, just like anything else, the internet

Jordan Wilson [00:12:39]:

has guardrails, you know, TV does as well. Yeah. Let's see what doctor Restafa has. Yeah. Yeah. Great. Great transition here. So, yeah, Doctor Restafa get us with a great question. saying there's 2 sides to ethics, the creative end and the user end. So kind of asking here what's the responsibility of companies like IBM who are, you know, actually helping to create. You you know, we haven't even talked about Watson x.ai, which we'll talk about here in a second, but What's kind of what do you think is the responsibility of the companies that are helping to develop these, AI systems? You know, obviously, on the back end as users, you know, we we have to have some of those ethics and guardrails, Moe, that you talked about in place, but what are your thoughts for, you know, companies that are helping create these AI systems?

How IBM approaches AI and ethics


Moe Alo [00:13:28]:

It's a huge responsibility. It's a great question. that's something that IBM, I'm very proud of the way we take it extremely seriously. We have a AI ethics board. We have a privacy board any application that is developed goes through a AI by design, AIFX by design process. We have, AI pillars of trust, transparency, privacy, fairness, explainability, robustness. any, and there you go. I love it. any, AI, involved development has to go through this process. of course, anything that touches on a military application, anything that comes close to surveillance spatial recognition. Those are things that IBM takes a very, very strong stance on. HR tools, you know, I think I read a a stat from, was a ZipRecruiter that I think almost 85% or something. a really high percentage. Don't hold me to it. of resumes are initially screened by a ai. And so think about the implications there. admissions into schools, education, political, right, political purposes, infrastructure, anything that's gonna be harmful. So and and this is we're a leader in this field. IBM, signed the role a call to action by the Vatican back in 2020 formed a partnership with the University of Notre Dame, with an AI ethics institute. there's been so many different regulations that have IBM has been behind and and regulatory bodies. And what what our position is is we believe in precision regulations. So we don't want to stunt the technology. Right? The use of the technology. That's where we want, you know, there needs to be guardrails. Right? But but we are proponents of advancing the technology always. That that's never something that we would be against, you know, at at least at this point in time.

Jordan Wilson [00:15:40]:

Yeah. And, you know, just just the piece like I love, Moe, that you broke down. like, step by step, what IBM is doing in regards to ethics in AI, because it is important, you know, for for end users or for those that are ultimately consuming content from other enterprise companies that are using AI. I think it's important to know the aspects here, that that you kind of laid out, but, a great a great question here from, Matthew. So we'll we'll we'll kind of transition here because you know, I think we talked about this on the show before, but as an example, you know, some big companies are cutting down or downsizing their ethics teams. Right? So Microsoft is 1. They laid off, kind of their their ethics team that taught, or that was, you know, working with ethics and AI. So what are your what are your thoughts on, you know, larger companies or, you know, we don't have to single them out, but you know, what are your thoughts on the the these companies that aren't maybe prioritizing ethics as highly as IBM or maybe as highly as they as they should be?

Why companies should prioritize ethics and AI


Moe Alo [00:16:46]:

Yeah. you know, I think everybody's learning. Right? IBM, we've been around for over a 100 years. We're one of the if not the oldest technology company. I would hope that there's a pivot there somewhere that it wasn't just, hey. Let's just take the guardrails off, and we're not doing Right? I would have to assume that there is something, that they have in place or in the works, but, you know, obviously, not something that I would that would be in favor that I would that would not be supported up, supportive of. I mean, we all know the the typical example or or let's say the easy example to understand is people that would be disadvantaged by an AI because of their ethnic group or or their zip code or they get a higher interest rate or they, you know, don't get approved for loans, etcetera. That's that's something that I hope we have a society really moved on and realized that that that needs to be taken care of. But the ethics, around AI are have to move as quickly as the AI itself is. Right? I know, there was a, a recent, study by I think it was by MIT Technology Review. and and this is a little more in the weeds, but it found that programmatic ads are spamming and and they're, you know, let's say Google's Google ads are similar. and they're being used on you know, websites that don't necessarily have correct information or might be misleading information. And so those are the dangers and then a lot of those are just generated by AI. So, you know, the market's being flooded with, people that wanna monetize. There will always be bad actors and we have to account for that.

What IBM is doing in the AI space


Jordan Wilson [00:18:28]:

Yeah. Absolutely. And, you know, one thing we haven't even talked about yet that I'm I'm excited. so let's let's quickly just kind of transition here because I'd love to talk real quick about what IBM is actually doing in the space because we we haven't even talked about this a lot, but, you know, Watson X dot ai. So, you know, most people, and we kinda talked about this to open the show. Most people have heard of Watson, and it may have been actually for society, you you know, seeing Watson, on, you know, jeopardy may have been people first big, or main introduction into what AI is, into what its capabilities are. But real quick, just talk about Watson dotai. so this is kind of more for enterprise companies, but, just just briefly motel us why Watson x.ai might be, worth looking at for for companies.

Moe Alo [00:19:20]:

Absolutely. And it's something that that we're very excited about. Like you said, IBM has a a deep history with AI and and Watson is a name that that everybody knows. And what we've done is we've created the Watson X platform and it's for enterprise, like you said. And it it it consists you have to think of it with with 3 subsets, Watson X dot ai, which is gonna validate tune deploy ai models. Watson X dot data allows you to scale the AI workloads for all data anywhere, you know, on prem, cloud, etcetera. And then watch the next dot governance and this goes to our topic today. You know, think of that as the nutrition label of your data. you know, responsible, transparent, AI workflows. you wanna know and make sure that the data that you're putting into the into your business safe, that is reliable, that is trusted. what we do is we help enterprises get off, to a good start with a solid, already trained model with where then the the customer would be able to insert their data and then also the domain or industry knowledge and that allows then the enterprise to have a structure. It has a at start, if you will, and then the guardrails that are necessary. So we're really excited about it. and obviously, it's it's off to a great start.

Importance of accessibility in generative AI systems


Jordan Wilson [00:20:49]:

Yeah. And and, you know, having access to these platforms, you know, whether it's you know, IBM's, you know, Watson X, and I know that they are even integrating, matters, llama too, into Watson X AI, but I think having access to these generative AI systems, whether it's at an enterprise level or just an individual, I think gives you so much extra benefit that maybe we didn't have a year or 2 ago. And I bring that up because, you know, Harold here joining the show has has a great question, because he's he's asking more about the approach to accessibility because, Moe, you kind of already, referenced this that, you know, this these generative tools aren't necessarily open to everyone. Right? Like, depending on, what what end, you know, because again, there's there's Google has their their tools and software Microsoft does. IBM does, you know, and depending on where you are or your budget, you may not have access to all of these. So what's what's the right way of maybe not asking from an IBM perspective, but as society, how do we tackle this accessibility piece and, you know, to kind of create a as level of a playing field as possible, because I think that's one of the biggest, One of the biggest issues facing generative AI use right now is this accessibility piece?

Moe Alo [00:22:15]:

It's a great question. And, Harold, thank you for asking IBM has always been extremely involved in the communities. Right? So we we have partnerships with universities with local schools. we have a whole division that is around accessibility. similar to what I said about the AI ethics, involvement in the design process, we also have a accessibility, check. And so now accessibility, you know, I I can see there might be 2 meetings access to technology, but also access for those who, have different abilities. Right? And so IBM has a focus on both. we launched a skills bill where we have a free online platform where people are able to go in, learn, earn certificates. and AI is obviously one of the the big topics in in that. And and I point you to skills bill.org. and there you go. There there you are. and so this is something IBM takes very, very seriously, and we're happy to, to do so.

Jordan Wilson [00:23:28]:

Yeah. And I know this is long show, but I do wanna end on this, Moe. You have such a great background, and I think you've shared so much great information today. But what would your what would your takeaway be? for, you know, business leaders for even individuals as we kind of grapple with this piece on on ethics and AI. What would your takeaway be for people that are trying to use it in a responsible way to to to grow their business? How can people do that?

Moe's final takeaway


Moe Alo [00:24:00]:

Well, I think they should, number 1, watch the everyday AI podcast every morning. but no. You do a great service. Jordan, I wanna make sure I get that, stated. I I appreciate what you do. You cut through and also highlight the things that are important. But I mean, as far as your question, I think what a lot of us don't realize or maybe haven't you know, thought about is AI has been used, and and everything that we do for a long time. Right? I think, what ChatGPT did was kinda shined the spotlight. You know, I think our CEO compared it to what NetScape did for the internet. and so which is great. which is, you know, something that everybody, is excited about and and the accessibility to to millions of people around the world to using, a a large language model like Chachi PT has been phenomenal, but my advice is to always ensure that you have ethics at the forefront There's a lot of damage that can be done to a lot of people and a lot of businesses, small businesses, if ethics is is not a serious component of anything, that you are developing or or, rolling out to the public. And so that that's really my advice. choose a great partner. you have in house talent as well as consultative, you know, you know, get a view of of the outside world of what other what your competition is doing make sure that you have, ethics at the forefront, and I think it's an exciting time. This is, you know, the one of the most impactful technological advancements, in history. We're all part of it. We're all, you know, right in the middle of it, and it's exciting.

Jordan Wilson [00:25:45]:

Yeah. Speaking of exciting, this, at least for me, and I hope if you're out there listening, in in tuning in live as well. I hope this was an exciting conversation because we went over a little bit of everything, Moe, from from ethics and the background of of even IBM to how we can, you know, look forward to responsibly using AI in the future. So thank you so much for spending some time on the everyday AI show to help us all figure this out together. Really appreciate it. I'll be watching. Thanks a lot, Jordan. Alright. And just as a reminder, don't worry if you couldn't keep up with everything, go to your everyday ai.com. sign up for our free daily newsletter, and we're gonna have a recap of the conversation that we just had with Moe, talking about ethics and everything else. And if you are listening on the podcast, don't worry. Look in the show notes today. We'll have links to this. And even if you wanna jump in, kind of our our daily LinkedIn thread and ask more questions, if if you are listening on the podcast and you're like, oh, it would have been great to join in live. Don't worry. You can click and do that. So, thanks again to Moe, for joining us. And thank you. And we hope you back tomorrow and every day on everyday ai. Thanks.

Moe Alo [00:26:54]:

Pleasure Jordan.

Gain Extra Insights With Our Newsletter

Sign up for our newsletter to get more in-depth content on AI