Introduction - Annemarie Penny
This post is part of our AI6P interview series on “AI, Software, and Wetware”. Our guests share their experiences with using AI, and how they feel about AI using their data and content.
This interview is available as an audio recording (embedded here in the post, and later in our AI6P external podcasts). This post includes the full, human-edited transcript.
Note: In this article series, “AI” means artificial intelligence and spans classical statistical methods, data analytics, machine learning, generative AI, and other non-generative AI. See this Glossary and “AI Fundamentals #01: What is Artificial Intelligence?” for reference.
Interview - Annemarie Penny
I’m delighted to welcome Annemarie Penny as our next guest for “AI, Software, and Wetware”. Annemarie, thank you so much for joining me today! Please tell us about yourself, who you are, and what you do.
My pleasure. I'm glad to be here. I am CEO and founder of a company called Dreamore. It's a newer name for a firm that I actually launched back in the 90s at the beginning of the dot-com boom in California. We are a recruiting and advisory firm for startups and middle-market clients. We have built thousands of startups and teams over the last 30 years.
And I'm an artist at heart, so I think that is ultimately where this comes from. I was pre-med with an art minor, which makes no sense. But I switched to business, broke my parents’ heart, and then came out to California and started working in this sector.
As a recruiting and advisory firm, we do targeted search and we leverage deep relationships. So it's a very specific kind of search that solves particular pain points that a lot of other firms may not be able to solve, or internally, sometimes you just don't have the time and the network to do that.
We are women- and diversity-led, so we particularly work with purpose-driven firms, companies that are building products and treating their people in ways that we would want to be treated ourselves.
That sounds great. So thank you for sharing that background. Can you tell us a little bit about your level of experience with AI or machine learning or analytics? I’m wondering if you've used it professionally or personally, or if you’ve studied the technology?
Definitely used it personally and professionally. I started my first job as a teenager at Hughes Aircraft Company and worked in the data department back when there were still some punch cards. If you've ever dropped a stack of punch cards on the floor as a teenager, and had to stop crying and get back to business - these are formative years! And so it is a perspective that machine learning and AI and tools like these have been around for a LOT longer than a lot of people realize.
And one of the reasons that I formed my recruiting company back in the 90s was to be in this dot-com thing before anybody else knew what that was. And so I draw some analogies to what I experienced in those early days with that kind of technology. The internet was so new that we didn't even have the imagination for what it was capable of. So a lot of it was putting HTML pages on a screen and saying, “Yay, that's the internet”.
So that's kind of where I feel we are with AI in a lot of ways. It's been around a long time and it's not mainstream yet. And we're only just beginning to understand the issues and the possibilities of a technology like this. Our tagline as a firm is “Bringing humanity back to hiring”. And I don't find that it's ironic that we created this newer tagline for our business, while at the same time implementing AI within our company. And so the spirit of that is, to the extent that we've learned AI up to this point, we're using it to do the things that we humans and our human brains weren't even formed to do.
Like I've trained my people for years that your career is not inbox management. Your life is not meant to be responding to texts as quickly as they come. And so I have very strong opinions based on experience that, if we can continue to work with AI, to work with technology in the ways that it was created and formed and is more natural, it's going to allow us as humans with our human brains to focus on things that WE can do. Like, AI is not going to tell your story. It's not going to relate to other people like we can relate to each other. So recognizing our strengths and weaknesses, instead of recognizing them as threats, or thinking they're going to solve all our problems - at both ends of that spectrum, that's not what it's going to be about.
But yes, my lengthy answer is just relating that, when the dot coms came, I popped off and said, “What are we going to need all this information for?” I had no imagination for how powerful the internet could be as a tool that we use constantly.
So you mentioned that you've used it both professionally and personally. Can you give some examples of how you've used it personally? And then also you mentioned that you're using it now in your company relating to hiring. So I'm really curious to hear about that.
Yes. So on the personal level, you know, with my artist’s soul, of course, when Night Cafe and DALL-E came out, you know, it was just fun to say “Draw me a futuristic McDonald's by the ocean” and then see what it would come up with.
So on a personal level, the early days, it was kind of more like a toy and a novelty. And occasionally creating some really poor writing on LinkedIn posts. And I decided to abandon that in favor of just using it as more of an editing tool. It's great for structuring thoughts, not necessarily good at telling an original story, in the way that I used it.
On the professional side, we are working with a firm that is creating some bespoke solutions with generative AI to solve some of our initial outreach. And that is very much an example where I finally was able to learn, “Oh, here's what AI can mean if used correctly.” I've trained a lot of account executives in my company, brand new recruiters, admins to work for me. And I'm drawing a very strong parallel in my experiences when you're training an LLM, “So here's my voice. Here's how to say this. And here's when to say it.”
And so, instead of just implementing it and wandering off, it's just like a brand new employee. If you hire an admin, and you don't tell them actually what you want from the job, and how they're going to meet their goals, and also forgive them when they screw up because, as a manager, sometimes you might have just not trained effectively. And instead of going, “Oh, ChatGPT just hallucinated, or it said something wrong”, sure, but which part of that is the fact that AI is learning from the humans? So if it's unethical, or racist, or biased, I'm not saying that's an easy problem to solve. I'm just saying, “Oh, if it's coming from me, then what can I do to be less racist and less biased and more ethical?” if that's where it's getting it from.
It's interesting that you've mentioned the racism and bias, because that's one thing that I've heard a lot from people that their concerns are with using AI in the hiring processes. On the one hand, there are some ways that it could help to mitigate bias. In other ways, it can reinforce existing biases that are there in society and in the data that's used for building those models and those tools. Curious to hear your thoughts about that!
You know, one of my early mentors in recruiting used to make a silly joke and he would say, “Oh, this job would be so easy if it weren't for people”, because we're literally in the people business. And so somewhere in bias and empathy and all these things, like, they're inherently flawed. We are still human beings interacting with technology and, you know, hoping that this technology can emulate humans in certain ways that it's ultimately coming from us.
So it feels like a bit of a Chinese finger trap, perhaps, where the more you pull at it, the tighter it feels. And… I don't have a pithy solution except to say, you know, 80-20 rule is a concept applied in many situations to recognize that there's no perfection. And you're going to do your best and be conscious about the decisions you're making.
So when it comes to recruiting, unless we have AI hiring bots, and there's no human being in the process whatsoever, we're still going to have that inherent bias in some way, and recognition is half the cure.
Can you share some specific stories on how you've used tools that included AI or machine learning features? You gave one example before about art, but can you share anything else relating to your personal use or your professional activities?
So we have trained an LLM to talk like me and think like me, to the extent that it can, just like hiring a sales rep. So this is where I had to do a little bit of soul searching, because there's all of this rhetoric about “AI is going to take our jobs”. And right in that same conversation is, “Well, we don't have lamplighters anymore, and we don't have people that type out Morse code and send telegrams”. And so as technology develops, without boring everybody with a whole bunch of statistics about whether this creates jobs or kills jobs, the jobs just evolve. And then we're evolving with that.
So what we have been able to do within my own firm, and I think this is where it's exciting for a lot of smaller firms that might be tight on budget, I can't currently afford to hire 25 sales reps in my organization. And so then it limits how much I can grow my firm, which by extension grows more jobs because of what we do for a living. So if I can take an LLM and train it, and have it working in the background - because I go onto my LinkedIn and I see it reaching out to people that I have very specifically said, “Reach out to these kinds of founders. Reach out to these kinds of venture capitalists. Reach out to very specific kinds of people. And say these very specific kinds of things. And then keep learning.”
Well, I also don't want to be staring at a computer 10 hours a day. So guess what? I've got a few hours of my day that I go be a human again. Because I can go take a coffee meeting with a friend or a client that I've never - I ran out of time because I got to stare at the screen all day. Or I can go out into the garden, which is so much where my soul is connected. And just breathe.
And so that is where it's just amazing that I can have this thing running in the background, instead of doing that very manual work myself, and then feeling overwhelmed by it, and then doing even less work because I'm just exhausted from it.
What are you finding as far as the responses to the outreach that your custom LLM is putting out there for you? It sounds like you’re using your custom LLM to initiate contacts with potential clients.
Right.
How are you finding that people are responding?
Very well. And this is where there's a lot of dynamics - just like hiring an employee. So what goes into hiring a person? And I'm going to equate it to basically training my LLM to do these person-like things. Well, first of all, you know, I have a 25-plus year reputation and a lot of existing contacts and authentic relationships that validates not only who I am and what I've been able to do in my sector, but also, how do I communicate and what's my reputation and all that stuff that had all that time to build up. And then knowing how to train the LLM just I would, as I would train an employee, “Call these people. Don't call these people. Say these kinds of things. Use a light touch.” We're not just selling. We're opening up a conversation that the whole point of that is to get them to me, so that we can actually like, “Okay, good, you're pre-qualified and there's a reason for us to talk.”
I kind of want to say another point, without going on too long, but part of the unfortunate reality of everybody's connected on Zoom now is that it is very easy to fill up your day talking to the wrong people all day. Because everybody's trying to connect and people aren't necessarily pre-qualifying.
So having a well-trained LLM to do that level of thinking, and then not bother people unless they're the sort of people that we actually do want to talk to, well, that's a nice lift in all directions.
So it's a very, very specific pain point that we're solving. And not having the unrealistic attitude that we're just going to push a button, and then it's going to solve all our problems, and then we're going to go to the beach now.
Getting that beach and garden time, that's important!
Yeah. So I guess the other thing with the AI is like, it's learning and we're learning. And it has no feelings, but I like to give it some forgiveness and be like, this relationship is new. You know, I've been married a long time. 10 years in, it's a different kind of communication than the early dating years. It's just our process.
Are there any examples where your custom LLM sent a message to a prospect that you looked at it later and said, “Oh, that's something I need to teach it to do differently”?
Sure. And because I've trained so many employees that are new in this business, that didn't off-put me, because everybody's going to screw up. And if you have the kind of management attitude and human being attitude that there is no mistake that is so horrible that you can't just take a beat, and go into solutions mode, and give some forgiveness in all directions. So I’m very much about mindset.
And there is a lot about the way that we're interacting with new technologies and interacting with each other in this remote environment we find ourselves in. You can decide that things are great or terrible. You can decide if you just ruined your reputation, or if you just made a mistake and it's learning. These are the same set of facts. It's just how you can interpret that, and then act accordingly, is going to inform the success. And the sanity!
Yeah. I'm curious if you include in your communications that are generated by your custom LLM that it was generated by an AI-based tool?
We spent a lot of time thinking about that and we've done some testing. And I can argue the point both ways. As a company that is known for transparency, our knee-jerk reaction was, “Hey, by the way, you know, we're getting this tool that's helping.” And the other side of that is, because we're in the recruiting and advisory business, there's a lot that comes with our role that has a lot of discretion, and understanding when do you filter information? Not because you're being duplicitous or sneaky, but everybody doesn't need to know everything all the time, especially if it's not adding any value.
So we decided for the early outreach - it moves so quickly - to just say, “Hey, do you want to get on a video and talk to me about what your pain points might be?” It moves very quickly through that. And I am setting up the lists of the people. So it is also very targeted. It's not just blindly wandering into the internet and talking to anybody. It is me behind saying, “Okay, can you just like do this?”
So there's generative AI involved. This is another thing that we're still in a phase that a lot of what is perceived as AI is actually just automation. And the generative AI comes in when it needs to be conversational and actually interact beyond that first stage. And maybe that's helpful for - people sometimes are like, “Oh, I don't want to use this AI”. And it's like, “Yeah, all you're doing is like, it's like an inbox filter.” It's just throwing things in different buckets so you don't have to. And even that is an uplift in your mental health and productivity.
So it's really in some ways more like having a template that you use and that it's simply filling in the blanks or customizing?
Well, there's a whole part of the LLM that - I just talked to it. “Here's my backstory. I'm from Arkansas. I lived in Israel. I like puppies.” You know, all that kind of stuff. So much of my personal brand and business brand is to relate to people at whatever level they're coming in. So if you were born in the South, we can talk about that. If you're from another country, we can talk about that, et cetera. If you're nerdy, if you're an artist, if you like to do physical sports, any of this stuff. So that is a lot of the stuff that I fed in the LLM, and just phrasing, just things that will make it seem natural.
So yeah, where we're testing this is: how much do we do beyond that? So we've, and this is the other thing about the hallucinating and the ethics and all of these other very, very legitimate concerns. You know, if you had a brand new employee, you're not going to automatically give them the company credit card with no limits and all of your passwords, right? Baby steps. And you might never do that, right? Because you don't know that this person needs that or that you can trust them.
So this was the big switch for me and whether or not it's helpful for others. Treating an LLM, because even the acronym LLM, it feels very science-y. And what does that mean, really? And I think that's off-putting for a lot of the mainstream is just all of these fears and lack of understanding. But if you figure, you know, your LLM is your new employee, and you got to train them. Or your new puppy. We can make it a puppy. Like, puppy's going to piddle on the rug. You don't want that. So you give them a mat.
Anybody who has been an early stage leader and a founder and went, “I'm going to hire people because I need help”, but then they don't realize it's going to take more before it takes less. There's an investment. And so going into implementing any kind of AI and recognizing: this is an investment. Even anybody that you hire, like, I have a swimming pool. So I just got a new pool guy. You got to talk to him about, like, “Here's how I use the pool” and this kind of stuff.
So I think that is where a lot of people are just getting disappointed in the results. Because it's like, if you hire somebody new, you're going to need to train them. And six months later is when you're really going to see the results. And if it's 90 days later, then you're lucky. That's just how it works.
Yep, that makes sense. So have you avoided using AI-based tools for some things or for anything in particular? Are there cases where you would choose not to use it? And if so, I'm curious about why.
100%. Two parts of the answer. Even though I've worked with so many innovative and tech forward clients over the years, I'm actually not the early, early, early adopter. I usually “let the line play a little bit”. So a year ago, I considered just making our firm solely focused on AI, because we've been so known for innovative technology.
I chose not to, not only because purpose-driven and female and diversity were topics a little bit closer to our heart in the grand scheme, but also, we have these challenges that I gotta make sure, like, “this thing isn't going to run off and steal all of our money and, you know, take our souls” kind of energy.
The things that I've avoided, it's a pretty clear line. Can I human this better than technology can? Or can technology do this better than humans can?
You know, 20, 25 years ago, we didn't have multiple inboxes and phones and texts and WhatsApp and Slack channels, da dat da dat da, right? You had an inbox, and somebody knocks on the door that you never answer, because who knocks on the door? And your phone would ring. That was pretty much it, to oversimplify.
So we've developed our technology to the level that last year I just kept muttering under my breath like a crazy person, “This isn't scalable. This isn't scalable.” You know? Because by the time you get through some of your LinkedIn, then WhatsApp is blowing up your phone, and now you got to go back through all this.
So where I've avoided using AI is when somebody is selling me some recruiting solution that is going to interview human beings and then deliver a report. And then presume that that goes straight to a hiring manager for interview.
And, granted, we are a boutique, so I see that these kind of tools can work on a much more volume, data-driven type of recruiting structure. But for what we do, it is never going to replace my ability to look at you across the room and just, like, sense you.
Okay. Yeah, that's a good example. Thank you for sharing that.
One concern that is common and growing to some extent is looking at where AI and machine learning systems get the data that they use for training their tools. And a lot of times these tools will use information that we’ve put into online systems or that we publish online. And the companies who build these tools are not always very transparent about how they intend to use our data when we sign up for these services or when we use them.
I'm wondering how you feel about companies that are using this data and content for training their systems. And what your thoughts are about these companies, what they need to do to be ethical. And whether they should be trying to get consent from the people whose information they're using, and whether they should compensate the people?
That's a lot to unpack. I feel icky. I'm not a lawyer. But yes, this is a very, very serious concern. And the fact that we don't even know the scale of that and the scope of that is an enormous concern. And it is across all functions. You know, we got artist friends and writer friends and musicians as well as, you know, just human beings.
So yes, whether it is through regulation or through other resources, I have much admiration and respect and support for anybody who is taking that on and making that their life's work, until we get something that is more palatable. I don't think that we're going to be 100% on that, but it's a huge concern, which also informs why we weren't really doing much with AI until it developed more.
And then the AI that we're doing is like a closed system. And those are the products that I end up liking. I have another client who came out of the mortgage industry. And there was just some government document that's a million pages. And so they built an AI system for the sector that is just accessing that labyrinthian document, but does not go out to the open internet. And then they can interact with it and do their jobs. More closed systems, the better, ideally.
Yep. So as someone who has used some of these AI-based tools, for instance, when you built your own custom LLM, do you feel like the tool providers, the companies who built the tools that you used, do you feel like they've been transparent with you or with us about where they got the data that they use for the tool? And whether or not it was trained ethically, and whether the creators were able to consent or be compensated for their content?
Very good question. We have talked about it. We have nerded out on topics. I think the short answer would be like, no, I didn't do a deep level of due diligence. And there does feel like a bit of “you don't know what you don't know” on that topic. So anyone who is generating more education around this without merely doing this for the purpose of inciting fear, that's useful.
Yeah so, my take on it is really that you shouldn't have to dig deep to find out that information, right? If they were being transparent, you wouldn't need to.
Yeah, I mean, you know, like we have certain systems that we rely on. People are going to drive in the correct lanes on the freeway and stop at stop signs. You really hope that we can get to a point that we can just assume that things are okay and not question. But I don't know. I don't know how we're going to get there from here.
Yep, it's definitely a tough problem.
So as consumers and as members of the public, our personal data has probably already been used by AI-based tools or systems. Do you know of any examples that you could share where it has affected you? (Without disclosing any personal information, of course.)
I wish I had a good one for that. I mean, we've all gotten, you know, hacked from this or that system and, you know, the “$2 class action lawsuit settlement later” kind of energy.
So one example is you maybe travel for your business. And so when you've gone through airports and TSA is taking biometrics and using machine learning to compare a photograph of your face to your identification.
Or social media where our data from, for instance, LinkedIn has been used for different purposes.
Or website signups where they ask for your birth date and then you end up getting things in the mail or through email, relating to assumptions about your age.
Oh, I'll walk carefully. I was in a mediation meeting recently. Nothing nefarious. But my representation was using the law to negotiate our side of it. And the opposing counsel had pulled information from the internet somehow and that was their line of reasoning.
(raised eyebrows reaction)
Yes, those were the eyebrows! So I made those same eyebrows while also recognizing - I said, that's very disappointing that my representation is using the law to discuss legalities and you, you're Googling me and pulling information from, I'm not sure where.
Just to add a little more flavor to it, very intentionally for many years, I've held back a certain amount from the internet. For good reason, right? And as we're scaling up Dreamore, it felt very important to communicate the story, you know. The Malibu Dream House burned in the Woolsey fire in 2018. It was very much on national news and, and featured. My company was in Forbes. There was a lot of things that's like, wow, I'm breaking and re-forming in certain ways. And transparency is the new secrecy. Let me show a much more authentic face to the world that I always thought I was, but it was still shielded.
And what has been very interesting is to be as open, not oversharing but just like, “Here's the story. Here's what we're creating. You know, let's all just drop our guard a little bit, and find ways to trust each other.”
And that scorpion tail has come back on me twice in the last month. And it hasn't been horrible. But it's almost felt like a test of like, where is that line? And there is a certain amount of my kind of processing where we used to make a joke like, “How do you know where the line is? And you look back over your shoulder like, oh, there it is, because you just crossed it.”
Yeah, it's been a very interesting moment to have this experience of not feeling scared, and to be okay to just say, “Hey, this is what's happening”. And then to have that come back at you so quickly is just, “Wow, how are we going to figure this out?”
It can't be that we either live in fear and we're shielded and we never do anything, or if we just try to be authentic, that somebody's going to try to harm us in some way. I mean, it sucks, but it does feel a bit binary. I have problems, no solutions in this answer.
Yeah, I think there's a lot of smart people trying to think of good solutions for it. And it's definitely been a challenge.
Generational too. I mean, you see where, you know, a few generations ago, anything goes. And now the internets are blowing up with the stuff that happened in the 90s and beyond. And you see some younger kids being a lot more guarded, or not on certain social media at all.
I'm not a Luddite. I love technology, but I am actually not complaining if we all just, you know, turn the devices off in the evening and go outside. You know, I grew up in a house where you couldn't watch TV over dinner. You had to actually talk to each other. Not the worst thing.
Yep. And back in our punch card days, everything we did didn't get recorded for posterity, right?
<laughter>
Okay. This is true. This is true. I do not miss - try to put those cards back together in the right order. Oh my gosh. If only there was a technology that could do that for us!
<laughter>
Yep. Well, final question and then whatever else you'd like to talk about, we can discuss. What we're seeing is that public distrust of the AI and tech companies has been growing. And to some extent, I think that's healthy because we're starting to realize just how much they're doing with our data and often without our consent. So I'm wondering what you think is THE most important thing that these companies could do or should do to earn and then keep our trust? And if you have any specific ideas on how they could do that?
So fear mongering is very popular and maybe it's just how our brains tend to work that, you know, everything is an ‘other’, and it's unknown, and therefore it scares us, you know? At our basic neuroscience level, you know, everything's either food or it's going to eat us. So we forget sometimes how hardwired we are, that technology, it's not as obvious.
Knowledge is power. I think the more - just my story about “transparency is the new secrecy”. If somehow somebody's got to throw the first diamond or whatever to just like, can we just open up the dialogue? And talk about it, and try to reduce the emotion around it, and the mansplaining and the fear and the lack of knowledge and all that? And just be like, “Hey, we're just trying to figure this out.”
That's very idealistic. I mean, I spent my teen years in the Middle East. I know that there's conversations that are so convoluted, and culture baked in, that it's hard to know where the out is. But we got to start somewhere. And having avoidance behavior and distrust, I think, sorry for the platitude, but “fear is the mind killer”. And so, in all directions, if we can just drop our guard a little bit, and have open conversations that are based in a certain amount of fact, that's going to help resolve these issues.
But it's not easy. Again, if there's something that I could say that would fix it, you know, that I'd probably be the multi-billion dollar CEO of whatever. And I'm not. We're all trying.
Yes, I think that's fair. We're all trying.
I've really enjoyed hearing your perspective on this and how you're using AI with your business. Is there anything else that you would like to share with our audience? Any upcoming events or anything that you'd like to let everybody know about?
Well, you know, Dreamore is scaling up. We're recruiting, an advisory for a small, medium startup businesses. We particularly help women and underserved groups and anything purpose-driven. So the one word on this is if anybody wants to talk to me, related to hiring. Got a great team, so it's not just me. At the end of the day, we love growing business. Every business is made of people. Even with all the AI inherent, there's still at least one person involved and we're your ‘people’ people to support all of that.
The only other thing I was going to mention on that last question is: critical thinking is such a wonderful skill. And I was taught that in school. I don't know that schools always mainstream this kind of analytical critical thinking. And so reacting to things with fear as opposed to like, it's okay to question everything and not make that a bitter cynical person. But just take a beat and question and then learn. If we can all do that, I think that we're all going to do better.
Awesome. Well, thank you so much, Annemarie, for making the time for the interview. I appreciate it.
Likewise. It's a pleasure.
Interview References and Links
Dreamore
Annemarie on LinkedIn
Subscribe to the Dreamore newsletter
About this interview series and newsletter
This post is part of our AI6P interview series on “AI, Software, and Wetware”. It showcases how real people around the world are using their wetware (brains and human intelligence) with AI-based software tools, or are being affected by AI.
And we’re all being affected by AI nowadays in our daily lives, perhaps more than we realize. For some examples, see post “But I Don’t Use AI”:
We want to hear from a diverse pool of people worldwide in a variety of roles. (No technical experience with AI is required.) If you’re interested in being a featured interview guest, anonymous or with credit, please check out our guest FAQ and get in touch!
6 'P's in AI Pods (AI6P) is a 100% reader-supported publication. (No ads, no affiliate links, no paywalls on new posts). All new posts are FREE to read and listen to. To automatically receive new AI6P posts and support our work, consider becoming a subscriber (it’s free)!
Series Credits and References
Audio Sound Effect from Pixabay
If you enjoyed this interview, my guest and I would love to have your support via a heart, share, restack, or Note! One-time tips or voluntary donations via paid subscription are always welcome and appreciated, too 😊
Share this post