Resources:
Join the discussion: Ask Jordan and Anatoly questions on AI bias
Upcoming Episodes: Check out the upcoming Everyday AI Livestream lineup
Connect with Jordan Wilson: LinkedIn Profile
Try Our Free AI Prompting Course: Register for our free Prime, Prompt, and Polish AI Course!
Navigating Cognitive Bias in AI: A Guide for Business Leaders
In the rapidly evolving world of AI, understanding the intricacies of cognitive bias is crucial for business leaders striving to leverage AI effectively. AI systems, particularly large language models, are powerful but not infallible. These systems, often reflections of the internet and human society, can inadvertently perpetuate biases that affect decision-making processes. This article delves into the nature of cognitive bias in AI and offers strategies for mitigating its effects in business environments.
Understanding Cognitive Bias in AI
Cognitive bias refers to irrational beliefs and perceptions that shape decision-making. These biases, while inherently part of human nature, are mirrored in AI systems due to their foundation on human-created data. Models trained on vast internet datasets may inherit skewed perceptions and biases embedded in that data. For instance, common biases like confirmation bias, where information that confirms pre-existing beliefs is favored, can heavily influence AI outputs.
The Role of Training Data
The training data of AI models plays a pivotal role in bias formation. Large language models, trained on extensive datasets, may reflect societal and internet-driven stereotypes. The core issue lies not just in data quantity but in how that data is labeled and prioritized. Human biases in labeling processes can unintentionally lead to skewed AI responses. Understanding this can help business leaders set realistic expectations and implement strategies to compensate for these biases.
Detecting and Mitigating Bias in AI Systems
For organizations relying on AI for strategic decisions, detecting and mitigating bias is essential. One approach is to audit the AI's outputs, examining them through a behavioral science lens to identify potential biases. Incorporating diverse perspectives during AI training and deployment can also reduce biases, ensuring more balanced and objective AI outputs. Moreover, engaging with AI in a 'trust but verify' manner, where outputs are critically examined before implementation, is crucial.
The Future of Bias in AI
As AI technology evolves, the anticipation is for models to offer better reasoning and decision-making capabilities. Narrow AI systems, designed for specific tasks, present an opportunity for reduced bias due to their focused datasets. However, the human element in AI development will always carry an inherent risk of bias. Continuous refinement and ethical guidelines in AI training are necessary steps toward minimizing these risks.
Conclusion
For business leaders, understanding and addressing cognitive bias in AI is not just a technical challenge but a strategic imperative. By acknowledging the limitations of AI and actively seeking to mitigate bias, businesses can make more informed and equitable decisions. Embracing this awareness is crucial for harnessing AI's full potential in growing a resilient and forward-thinking organization.
Topics Covered in This Episode
1. Understanding Cognitive Bias
2. Cognitive Bias in AI Models
3. Training Data and Model Development
4. Future of AI and Managing Bias
Podcast Transcript
Jordan Wilson [00:00:17]:
Do you ever just blindly copy and paste what a large language model gives you? Right? I get it. We're all overworked. We're stressed. There's so many things. Your, you know, your manager is demanding more now that you're using AI, but that can actually be very dangerous. Right? Just blindly trusting, what a large language model like ChatGPT or Gemini or Copilot or Claude spits out. And one of the biggest reasons, and I think, a reason that sometimes AI fails is because of bias. Right? Essentially, large language models are a reflection of the Internet.
Jordan Wilson [00:00:56]:
They're a reflection of society, and there's a lot of things wrong and sometimes these models aren't the absolute truth. Sometimes they're very flawed. So we're gonna be talking about that more in-depth today as well as what you can do about it and how to keep an eye, for different types of bias biases. Right? I guess it's how it's said. That can show up in your large language models. Alright. I'm excited for today's conversation. I hope you are too.
Jordan Wilson [00:01:22]:
Welcome to Everyday AI. Maybe it's your first time here. If so, where you been for the last three years? We do this every single day. My name is Jordan Wilson, and, this is your daily livestream podcast and free daily newsletter helping everyday people like you and me not just learn AI, but how we can leverage it to grow our companies and to grow our careers. I want you to be the smartest person in AI in your department at your company. So if that's what you're trying to do, you're gonna wanna go to our website. That's youreverydayai.com. We're gonna be recapping, today's conversation, as well as really recapping everything else you need in the world of AI, and we do that every single day in our free daily newsletter.
Jordan Wilson [00:02:01]:
So if you want more insights from today's show, make sure you go sign up. Alright. Before we get started, and I'm excited to talk about the top reason for AI failure, cognitive bias, let's first go over what's happening in the world of AI news. So Microsoft has launched two new AI sales tools, and it looks like it's to compete directly with Salesforce. Yeah. Salesforce, kind of picked a fight with Microsoft, and Microsoft has now introduced two new AI tools, a sales agent and a sales chat to streamline their sales process as part of its Microsoft three sixty five Copilot platform. So Sales Agent automates lead qualification, meeting scheduling, and follow ups, while Sales Chat delivers actionable sales insights using CRM records, emails, and meeting notes. So both tools integrate with Microsoft Dynamics three sixty five and Salesforce, funny enough, minimizing reliance on traditional CRM systems.
Jordan Wilson [00:02:57]:
So this tool will be available in public preview in May, signaling Microsoft's aggressive expansion into AI powered business applications. So, yeah, Salesforce's CEO had has been a little critical of Microsoft's AI approach and, you know, they launched their agent force. So, Microsoft just clapping back. Alright. More big tech. Google has launched its new AI mode in search. Kind of going after perplexity and ChatGPT which were eventually or which were originally just going after Google. It's like this full circle weird moment.
Jordan Wilson [00:03:29]:
So anyways, the new AI mode is part of the Google One AI premium subscription plan, and you can access AI mode starting this week through Search Labs, which is Google's experimental platform. So the feature is powered by Gemini two point o, Google's latest AI model, which enhances reasoning, thinking, and multimodal capabilities to handle exploratory and comparative questions effectively. So AI mode uses a query fan out technique to issue multiple related searches simultaneously across, various data sources, consolidating results into detailed and accurate responses. So, yeah, you can if you are a paid subscriber, you can access AI mode or at least sign up, on on the wait list, by visiting search labs or you can go to google.com/AImode. Alright. Last but not least in AI news, a big, big one. OpenAI is betting people are gonna really love agents so much so that they reportedly might be offering one that costs $20,000 a month. Yeah.
Jordan Wilson [00:04:32]:
That's a month. So OpenAI is making headlines with its bold pricing strategy for its advanced AI agents, reportedly charging up to $20,000 per month for enterprise level automation tools. So these AI agents are described as PhD level and are designed to take actions on behalf of users targeting large companies looking for scalable automation solutions. So a lower tier version of the AI assistant, might be priced at $2,000 a month, and that's aimed at high income professionals seeking premium AI capabilities. This marks a major shift from OpenAI's previous subscription model where its highest price plan was $200 a month. So this is not, you know, this is just according to reports. These are more or less just well vetted rumors. So this hasn't happened yet.
Jordan Wilson [00:05:18]:
So, you know, don't don't log in yet to your ChatGPT account, you know, trying to sign up for a $20,000 I mean, that's wild. Right? Alright. Enough about that. We have more on those stories, and a lot more in our newsletter. So make sure you go to your everyday a I Com and check it out. Alright. But, let's talk cognitive bias because I think so many people are just blindly following what comes out of a model, not even knowing some of the dangers, that you might, that that entails. So, please help me welcome to the show.
Jordan Wilson [00:05:49]:
I'm excited for today's conversation. I hope you are too. So we have with us Anatoly Shilman, the CEO and cofounder of Cod Bias AI. Thank you so much for joining the Everyday AI Show.
Anatoly Shilman [00:06:00]:
Thanks, John, for having me.
Jordan Wilson [00:06:01]:
Alright. I'm excited for this one. Live stream audience, thanks for tuning in. Big Bogey and Michelle and Marie and Jamie and Vincent and everyone else. If you have questions, get them in now. But, let's start at the top, Anatoly. So what what is cog bias? What is it that you all do?
Anatoly Shilman [00:06:16]:
Well, we we built a platform that detects and mitigates cognitive bias in communications. We what we what started off as our own internal project to help us ask better questions during customer discovery is now to turned into a full platform. We're able to take people's questions for customer discovery, marketing research, NPS scores assessments, anything, and, give them a breakdown of the biases they may be facing within the questions they have. And, also, on top of everything else, we rephrase and give them better suggestions on how to do it better. The same now applies for their emails. So if you have that difficult email to write, say, gotta break some bad news or an angry mail, you know how they say wait twenty four hours to write an email. With our client, you can actually do write that email. And our system, based on the, based on the actual context you give it, we'll rewrite it for you in a better way and also tell you what was problematic about your original email.
Anatoly Shilman [00:07:08]:
And we found that a lot of folks such as salespeople, obviously, marketing research, UX, UI, product managers have been using our product. And, it's funny. Well, one of the things we've been discovering that people are constantly creating new methods of using it. One of the things that's coming out very soon, as you mentioned, AI agents, is that we're we're actually able to audit AI agent conversations and detect the biases that they have and and make go make reports for companies to make the changes necessary to make them better.
Jordan Wilson [00:07:38]:
Mhmm. So I I wanna kinda skip to the end here, and this is kinda how I started off the show. Right? Because I think so many people just blindly either copy and paste what comes out of a large language model or they just inherently trust it, as as being accurate and, factual and, you know, bias free. Why is that a mistake?
Anatoly Shilman [00:08:01]:
Well, the biggest is because, just like, you know, AI is like people. It was built by people just like Google was built by people. You know, before we had the whole AI explosion and people went on Google. Well, it's on Google. It must be true. Was, like, a very constant refrain that people gave, and that's just absolutely not correct. And, I think one of the things that you'll discover in AI more and more often, we call people are calling hallucination. It's really more of a form of BS.
Anatoly Shilman [00:08:27]:
There's actually a very popular paper called Chegg GPT is BS, and it was written simply from the perspective that, you know, AI is really more like the your know it all friend. You everybody has one in their circle. They tell you all these incredible things, and most people are like, wow. They know everything. They expect accept it as fact. But what the AI has to do is they have to answer your question. So even if they can't find the answers, they'll make the answer up. Mhmm.
Anatoly Shilman [00:08:50]:
Obviously, many lessons in that. Most recently, one of the bigger law firms in America had a scandal where they, that was being used to for them to, you know, do casework with actually created precedent cases on its own. So it created a whole universe of the cases that never existed. And it's a consistent theme over and over again. What what we end up with is, you know, people are so enthusiastic about this new leap in innovation that they forget that just like anything else, you should have an attitude of trust but verify. I recently did a TEDx event, and, the conversation there was, you know, the future we make and the future that is AI. And what was most interesting about it is the initial when I asked people, how many of you are enthusiastic and excited about AI? You know, like, the whole room raised their hands. It was very exciting.
Anatoly Shilman [00:09:38]:
But as soon as I started talking about some of the things that have been observed, and I actually it started be asking people individually, like, well, tell me your real thoughts on AI. It was always a trust but verify attitude. And I think the issue is a lot of time people think, well, you know, obviously, look. It's made by OpenAI. It's made by Microsoft. It's made by Google. It has to be good. But they're just like they're not infallible.
Anatoly Shilman [00:10:01]:
They can make the same mistakes. And because they're built by engineers, they have the biases that those engineers have. So they have the same human extensions of our personalities. So the way they gather the information, the way they disseminate it is reflective of humanity.
Jordan Wilson [00:10:15]:
So I think maybe let's break this down, piece by piece or, you know, we'll we'll go chain of thought on this episode title here. Right? But, you you know, what what is cognitive bias? Right? I think all people kind of understand what it is, but what is actual cognitive bias?
Anatoly Shilman [00:10:30]:
The best way to kinda think about cognitive bias, you know, there's a really long scientific definition that I will not bore you with. It's honestly irrational beliefs based on our perception. That's the best way to kind of extremely simplistic mind. Just so please, none of the psychologists in the crowd yell at me, but I'm just trying to make sure that it's something easy to understand. And when I say that irrational is that we don't think clearly when we have certain beliefs. Right? We try because our ability to have cognitive biases is what gets us through the day. Ultimately, we make a choice every morning when we get up to get dressed a certain way, to do our hair a certain way, to drive a certain type of car and everything else because of the way either we wanna be perceived or we perceive ourselves or the feeling that it gives us. These are all biases.
Anatoly Shilman [00:11:13]:
The key point of cognitive biases, they're not bad. They're just part of our humanity. So in some cases, you know, most some of the most well known biases are there confirmation bias, break framing bias, availability heuristic. Those are things that help us and hurt us whenever the situation calls for it. You know, sometimes availability heuristic is reach for the first thing that is closest available us to us to solve the problem that we have. So in some cases, it's a hammer to nail, to nail a a a a nail to the wall. Other times, it's going to be a flat object because that's the closest thing to us. It's kinda the same day we operate with a lot of the things that we do.
Anatoly Shilman [00:11:52]:
So choices from a perspective of, hey. I need to get in I have to send out a survey to my customers. Let's ask ChattGPT for the top 10 questions about car buying. ChattGPT spits out the questions, and bam. All of a sudden, you get through a whole process where it becomes, you know, here's the questions. And you're like, well, they sound good to me. They're perfect. Mhmm.
Anatoly Shilman [00:12:14]:
There's no breakdown. There's no analysis. There's no belief.
Jordan Wilson [00:12:17]:
So I I I wanna break down two keywords I heard you say there. So, you know, you said irrational beliefs based on perception. So beliefs and perceptions. Right? Because these are things that most people probably don't think go into large language models. Right? Beliefs and perceptions. Those aren't fact based. Those aren't scientifically research. How does that happen, and how can people be on the lookout for when that does come through a large language model?
Anatoly Shilman [00:12:48]:
Well, I wish there was a simple way. Right? And the first thing is it happens because humans are the ones who make it. So even if AI makes another AI, it's based on the original programming of the human. So you're just going to have a new permutation of the same biases or an evolution of some new biases based on the old biases. The key thing to understand is, like, you know, for instance, an engineer will program an AI and say, I want you to put the information out this way. And I want you to put when they ask for a list, this is how the list will be based on this thinking. And, when you're pulling from, news sources or media sources, this is the first five hundred you're going to look at before you look at anything else to solve the problem. Is it because of their personal beliefs? Is it because of, their perception of what's reputable versus what's not? As a new source, we don't know.
Anatoly Shilman [00:13:34]:
Right? And the it's the same applies the same applies on the other end. As it kinda goes through the process of of coming up with answers, if it can't find it in those 500 and I'm just making up that number. I don't really know what the real secret sauce is in those cases. All of a sudden, it becomes a situation, well, they're gonna if they can't make find in those 500, it may think the other ones are less reputable. So instead, it'll come up with its own answer.
Jordan Wilson [00:13:58]:
Mhmm.
Anatoly Shilman [00:13:58]:
Or it will add its own little spin to it. And because it's a AI, you think, well, it's a computer that answered it. Must be correct.
Jordan Wilson [00:14:04]:
Yeah. Yeah. That's that's the worst thing you can do with a large language model. Right? Is this like, oh, it's a computer. It has to be right. But so many of the things that we ask large language models, there's nuance. Right? It's not binary. We're we're asking it for strategy to make decisions.
Jordan Wilson [00:14:18]:
We're not necessarily always asking it to count the number of r's in Strawberry or, you know, the capital of of Illinois. Right? But maybe if you could, could you walk us through just what are the types of of biases? And and, you know, maybe just briefly, you know, like, I know, like, you know, confirmation bias. Right? Maybe could you walk us through briefly, you know, two or three of the most common types of of biases that that show up in large language models and and what they mean?
Anatoly Shilman [00:14:44]:
Yeah. Obviously, confirmation bias is probably the most well known one. You know? It's confirming in it's in its own initial beliefs. So quite often, what it'll do is the way it best it's best to consider is not from the point of the AI. It's from the point of view. How biases really impact us is our perception of what is being said to us, shown to us, etcetera. So, AI is going to respond to us in a specific way and bias us in that way. So in some cases, we'll ask it a certain question.
Anatoly Shilman [00:15:11]:
It'll respond back, and it'll trigger our confirmation bias because it's gonna be confirming our facts. Mhmm. So if we ask an AI question that has a obvious answer, it's going to spit it back out at us in a specific way, just make it, you know, prettier, more or less, or more sophisticated or expand on it more. So confirmation is a big one. Framing bias. We frame something in a specific way to get a specific answer back. So if we say just make it up. Mercedes Benz is the fastest car and the best car for the money based on the luxury blah blah blah, and then we're gonna ask questions about it, now the AI is going to be responding back in the same way just like humans would.
Anatoly Shilman [00:15:48]:
Because, again, the AI is not here to argue with you. I know we've seen those comical stories where AI starts arguing facts with you, but that's not really the reality how it operates. And then, obviously, I talked about availability heuristic, which is, which is, one of the most, interesting ones because, like I said, it's the lowest hanging fruit. Mhmm. And yep.
Jordan Wilson [00:16:14]:
You you bring up a a a fascinating, you know, point here that I wanna dive a little bit deeper in. Right? So when when models are, you know, essentially mirroring our our own beliefs. Right? But I think what's important to call out is, you know, a system prompt. Right? All large language models have system prompts. And and one thing you said there is most of them, they are designed to be a helpful assistant. Right? So, even if there's not an answer, they kinda wanna be helpful. And and I think that's why sometimes you get these, you know, halfway answers or, you know, things that maybe you you look at and you're like, is this right? Well, sometimes it doesn't always matter because it's ultimately trying to be helpful. But, you know, I wanna ask you, how does the conversation in the context of, of a large language model when we're, you know, whether it's Copilot or ChatGPT or whatever, how is that going to influence it? Like, actually how we're prompting it and and, you know, what the outputs we get in terms of bias?
Anatoly Shilman [00:17:11]:
Well, it's it's a huge it's actually a big, big factor. If you think about it from a perspective, say you ask it for for a specific element or specific answer, and then you say, well, now I want you to write it, but pretend you're a twenty year experienced engineer and write in that number. Write it nicer. So now you where oh, there's the bias element. What is truly a 20 engineer? How do you write nicer? What is nicer? You know, sheer definition elements and how the biases are perceived from then becomes a hot mess. And quite often, that's where the prompting kinda falls apart. And that's why for a while, people are like, well, you know, we don't have to know how to do prompting anymore because AI is so smart. I'm like, unfortunately, we do because the one thing that AI claims to do that it actually doesn't do is it doesn't really understand well.
Anatoly Shilman [00:17:58]:
It understands, in a very initially, in a very bare bones thing. I heard a very great quote yesterday at an event where they said this. They said, at this point in time, AI is the worst it's ever going to be. And, that's a it's a true statement. Right now, we're at the very beginning, at the very earliest stages. So quite often, people have expectations of a flying ship when we're probably somewhere closer to a horse drawn carriage by AI standards. We'll get there, but the problem is as the more things get sophisticated, the more complex they'll get from a perspective. How do we ask a question that is perceived by the AI in the right way? Because we'll say write it nicer.
Anatoly Shilman [00:18:36]:
So it'll change a few words. It'll sound nicer to us. But if the context, is still the is still something that has harmful biases in it to what our objective is, it was not really helpful to us. No. It was just an answer given because I write it nice. So fine. I'll put some puffery around it. I made it nicer.
Anatoly Shilman [00:18:54]:
And, so our prompting doesn't necessarily help it be better at its job. Our prompting just helps it, again, confirmation bias, helps us confirm that, helps it confirm that we want something nice as the road changes stone to its perception of nicer, but not necessarily solve the actual problem.
Jordan Wilson [00:19:10]:
Let's maybe talk about the root of this. Right? Because, you know, I kind of I I I kind of reference that, you know, large language models are a reflection of the Internet and that's a reflection of of humanity and right? And that's why there's sometimes stereotypes and and biases, you know, to begin with. But walk us through how are models, like, actually reflecting these biases in the long run? So maybe can you just walk us through training data? And, like, where do some of the, you know, issues in terms of cognitive bias, where do they get inserted into this whole equation when it comes to training data?
Anatoly Shilman [00:19:46]:
Well, I mean, you, right, we kinda talked about the idea when they even begin the telling the AI, this is how you're going to, parse out information. This is how you're gonna pull it apart. This is how you, based on these requests, this is how you're going together together. Then we have to start thinking again. Once those, issues are inserted and once it, the AI has to start thinking, what's a repeatable information source of information? And then it starts, trying to pull that data out. We still have to consider the element that it's very much the equivalent of drinking from a fire hose. You know? I can't even imagine the sheer amount of petaflops and god knows what other measurements we could apply of information that are constantly flowing through that it has to parse through to the to delineate whether or not the one specific minuscule thing that we're asking has the answer for it. You know? And it still returns it in seconds.
Anatoly Shilman [00:20:39]:
And so that stuff is where you really start, falling apart on training data because it's not you know, there was recently a big thing with DeepSeek, you know, and how they trained it for 5,600,000. Obviously, it's not true. But it was a cute
Jordan Wilson [00:20:53]:
a whole episode on that, but thank you for calling that out.
Anatoly Shilman [00:20:56]:
Yeah. It was it was a cute it was a cute number, though. Right? And, but the big thing about it is that kinda brought to the forefront, what does training really mean? Mhmm. Right? And And when we think about training a small model, it's literally us humans sitting down and working through the labeling elements of specific data points and how the AI should treat those, elements based on the labels we assign to it. Mhmm. In the case of a massive model with massive, massive, massive amounts of data, it labels it itself. So it's only thought to label things. And what's the issue when you attach AI to label things? It, you wish it was more like, you know, all colors of the rainbow being able to see all points.
Anatoly Shilman [00:21:39]:
It can. It's much more limited. It doesn't have that neural depth that we have right now. So instead, what what it does is, you know, it's more or less, if and then rules. A lot of those are applied. And I'm simplifying way too much. That's not really how AI is. But for the basis of understanding, it's really how it kinda perceives, information.
Anatoly Shilman [00:21:59]:
Does this answer the question? Yes? No? Next. Does this answer the information? Yes? No? Next. And it goes through that whole routine. And when it gets to a certain point where it was like, well, this partially kinda answers the information. If you extrapolate this and then kinda, you know, smooth it out, which is your BS factor, that's the information. Yeah. So so in every hallucination or BS, point that AI makes, there's always elements of the truth in it, which makes it so convincing. And that that's the biggest thing to consider, like, during the training.
Anatoly Shilman [00:22:29]:
Because when you consider what's the training, you know, they do spend months and months of potentially years training models, but it's not like they're hand labeling stuff. What they're really doing is just overseeing this enormous model trying to label everything. And then doing audits and checking and checking again and rechecking and saying like, woah. That's way wrong. You know? And if it's big enough, they catch it. The problem is sheer amounts of data is just not possible right now to catch everything. But whatever be, I can't possibly answer. I don't think they can either.
Anatoly Shilman [00:22:56]:
Yeah. And that's why even though you see these new evolutions in, you know, we saw these new evolutions in agents. Literally, almost every week, a new tool comes out, and it feels like it's the next tool and the next tool. The consistent theme is the same. They're not actually creating better depth. They're creating better response time, maybe a lower latency, cheaper. They're sometimes making a better conversational piece to it, But the info being put out is still very much the same. And for us, when we actually look at the idea how it measures cognitive biases versus the scientific models we have, the consistency of Strad GPT and Claude and a couple of the others were around 30 to 40% versus what we do.
Anatoly Shilman [00:23:34]:
And the reason being is because, again, sheer amounts of data and how you label it and how the scientific application actually applies to the, to a specific word or specific nuance in the sentence is not the forte. So when you see, like, where AI is heading, we have the general AIs, and I think you probably might have talked about this already in the past, the explosion of narrow AIs Mhmm. That are going to be good in specific elements, and that's going to be their main motif.
Jordan Wilson [00:24:05]:
Are you still running in circles trying to figure out how to actually grow your business with AI? Maybe your company has been tinkering with large language models for a year or more, but can't really get traction to find ROI on GenAI. Hey. This is Jordan Wilson, host of this very podcast. Companies like Adobe, Microsoft, and NVIDIA have partnered with us because they trust our expertise in educating the masses around generative AI to get ahead. And some of the most innovative companies in the country hire us to help with their AI strategy and to train hundreds of their employees on how to use GenAI. So whether you're looking for ChatGPT training for thousands or just need help building your front end AI strategy, you can partner with us too, just like some of the biggest companies in the world do. Go to your everydayai.com/partner to get in contact with our team, or you can just click on the partner section of our website. We'll help you stop running in those AI circles and help get your team ahead and build a straight path to ROI on GenAI.
Jordan Wilson [00:25:09]:
So a great a great question here, from our audience and you kinda mentioned about the future of AI. Right? Like everything going, you know, agentic or or multi, you know, multi agent environments. But, you know, obviously, now we have these models that, you know, think, these models that reason, these models that, you know, take their time. So good question here from Cecilia asking, you know, what do you do to detect cognitive bias and encourage maybe slow thinking versus the fast thinking when we want AI to be fast? Yeah. So, like, I'll even add on to her question. You know, do reasoning models, you know, that take their time to think, is that also a a process that maybe we're gonna see less bias?
Anatoly Shilman [00:25:51]:
We may potentially, but, again, it still falls back on the developers. Right? It's a unfortunately, it's like a wheel. As soon as you insert humanity into the wheel, we have our biases. All of us do. That's not a bad thing, like I said. It's just, unfortunately, how we perceive information and other things, will dictate some of the more harmful biases that pop up, which is quite often when you hear about stereotypes and stuff. It's not that the engineer wrote, yes. Men are better than women or some other thing.
Anatoly Shilman [00:26:20]:
No. It's literally their biases and how information is parsed, what gets priority over what causes the model to extrapolate into the next piece, which is, well, this is how I'm going to perceive things. And while they have hundreds upon hundreds upon hundreds of engineers constantly looking and auditing and checking, it's just, again, such a vast amount of information and such a vast amount of variance that that's not not possible. To Cecilia's question, what you're dealing with, you know, there's a great book by Danielle Kahneman called Thinking Fast and Slow, which kinda is, I don't wanna say he's the father of cognitive modern cognitive bias behavioral economics. He kinda is, though. Right? And so the best way to kinda consider it, there will be a space for these models that are more analytical and focused on things. It's the same thing as we have other things that do certain more complex things. I always liken it to the idea.
Anatoly Shilman [00:27:10]:
You know, people have an Apple Watch and then people have a manual watch that still has gears in it. And they prefer that element because they feel to them it's more reliable because it doesn't run out of batteries.
Jordan Wilson [00:27:21]:
Yeah. It
Anatoly Shilman [00:27:21]:
just moves because you move.
Jordan Wilson [00:27:23]:
Yeah. I think I I think that's a great example, and, you you know, we'll definitely put that in the newsletter just kind of about the, you know, system one versus system two thinking. I think it's important, you know, when we think about, you you know, using AI. Another good question here from Douglas asking, are there some models that have more inherent bias than others because of how they were trained?
Anatoly Shilman [00:27:45]:
Honestly, that's very subjective. Unfortunately, all of them have bias, and there's no such thing as harmful bias. It's simply their perception of certain information points or how you ask questions. Right? What quite often, what we do is that what we noticed with a lot of our users is they actually will generate stuff on, like, ChatGPT, dovetail, name a ma name name a system that generates questions, and then they'll run them through our system. And then that that that will be the result that they use for the actual, for their actual, product. And the reason why they do that is just twofold. Sometimes they're fine with the questions. They just wanna understand, how am I going to be biasing my audience? Because no matter what you ask, you're always biasing them some way.
Anatoly Shilman [00:28:23]:
But the question becomes, am I biasing them just to answer the question truthfully if I'm doing marketing research? Or if I'm doing sales and I want to prompt them to action, am I doing that effectively? So those are the things you're really pushing on. But right now, for the most point, like, for instance, you know, I think if you look at size of model, large language models have a lot more inherent biases than smaller models. And that's just sheer by the sheer amount of data they consume and also the sheer amount of hands that touch the model. At the same point, narrow AIs totally have a bias in them. And the key element that we try to do you know, I, I'm always been a big proponent of having a very diverse team look at the model and label it and look at everything else and how it perceives data. And that's the reason why. Because if you have a a large variety of people look through it, you're not going to have the perspective on how to parse data from one or two individuals or 10 individuals. You're going to be able to do these things in a much more, you know, almost, objective way to a degree if I if if it was the proper word.
Anatoly Shilman [00:29:25]:
And when we developed our model, for instance, and, again, narrow AI. Right? Very different. We developed we started with 17,000 questions, and we looked at their sentence structure. And from them, we're able to extrapolate to 450,000 questions. It's somewhere in that neighborhood, probably even more. It's the one of the numbers I heard. And the idea is now we're able to label those 17,000. We can give the model the nod of these are the scientific facts assigned in those cognitive biases.
Anatoly Shilman [00:29:52]:
This is how you're going to react to this based on the deaf scientific definition, not our perceptions, but what science defines. Now does that mean science does not have biases? No. But science based on the best thinking we have in existence for our society, this is what it is. And as it evolves, we evolve. This is how the bigger models do too. Because as they as they keep coming out with new versions, it is our hope that they keep updating that particular element of the of how their model thinks. But the the question I have, and I think everybody has, is, you know, with such a fast evolution cycle I mean, we're talking sometimes, thirty days between new releases, not if less. Obviously, a lot of that stuff is maybe quick fixes, but you always have to wonder what's going on in the background.
Anatoly Shilman [00:30:36]:
Are they actually fixing the main issues, or are they just adding to them by creating more features?
Jordan Wilson [00:30:42]:
Yeah. Yeah. And I think sometimes you you you spend time, you know, circumventing, some of those shortcomings, and then new model comes out and it's like, okay. What about all that work that we put in, you know, to build bridges, you know, around over or through some of these, you know, inherent problems? But, you you know, we talked about a lot in today's conversation, you know, from everything from, you know, bias and training data and and, you know, bias in the humans that are building it, the types of, cognitive bias, which is super helpful. But as we wrap up here, I think I'm gonna, toss this over to, Big Bogey here on YouTube. I think this is a great way to wrap the show. So he's asking, how do you remove unwanted bias when that bias may be deeply entangled with essential data? And I'll even say, you know, hey. Aside from using your, you know, your platform, how should companies be tackling this? Because it's a huge issue.
Anatoly Shilman [00:31:33]:
Well, I think the element has to be is, the value of the data and the value of the output of the data. Right? Because quite often, if you're looking at specific data where the bias may, for instance, affect the key decision the company is making, getting outside help will be essential. Because, and, obviously, my my platform can help with a lot of the question stuff and the other stuff. But when you're looking at the bigger pieces, like, there's consultancies like Precipio out in out in California whose entire stick is to look at cognitive bias and how it impacts decisions. Because, the elements that he's talking about you know, essential data may have, cognitive biases deeply entangled in it, but, ultimately, our perception of those of that data is what causes the biases that will impact our decision making. The data will bias us in a way, but it we have to make the choice of how we receive it. And if we're already aware that we may have a problem with it, that's when we have to seek outside counsel on it and be able to almost bring a a pair of eyes to oversee our process and to figure out if we need to have a more object a subjective process or objective process to how we make choices. Because, ultimately, again, humans.
Anatoly Shilman [00:32:43]:
Right? Trust just like trusting a machine, well, it's not a great idea, especially when the data is complex or maybe, to a degree, maybe very very much human related and something that an AI could not possibly comprehend the same way, then it becomes a situation. You have to make your best choices. There's a lot of good classes, a lot of good reading, a lot of good curriculum. I've always been a huge proponent of companies doing behavioral science and behavioral economics training for the very reason, not because it can replace a tools like mine, but it can enhance people's ability to be able to spot the issue and then, seek the solution versus, you know, finding that after the fact, oh my god, our sales calls and our sales meetings and the the way we extrapolated this data from these sales numbers was completely wrong. The customer didn't really want this. They just felt they had no choice until they found a better solution.
Jordan Wilson [00:33:35]:
So much good, advice there. I I I think, Anatoly, thank you, so much for taking time out of your day to join the show. You helped us, I think, make much better sense out of a very complex and a very important topic that we all need to understand. So thank you much. Thank you so much for your time and insights.
Anatoly Shilman [00:33:54]:
Thank you so much for having me, and, I really appreciate the podcast. Honestly, I enjoy it quite a bit.
Jordan Wilson [00:33:59]:
That's great. Hey. I do too, but, you know, it'd be weird if I didn't. So, hey. If if you enjoy the podcast too, if you heard something here from Anatolyan, you know, like, wait. What was that? Don't worry. We're gonna be recapping it all in our free daily newsletter. So if you haven't already, please go to youreverydayai.com.
Jordan Wilson [00:34:16]:
Sign up for that free daily newsletter. If this was helpful, tell someone about it. Please subscribe to the channel. Leave us a rating. Tell your friend. Tell your mom. Tell your neighbor. Tell your mom's friend's neighbor.
Jordan Wilson [00:34:26]:
But more than anything, make sure to join us tomorrow in everyday for more everyday AI. Thanks, y'all.
Anatoly Shilman [00:34:32]:
Thanks.