In this Forbes Talks interview, Forbes Associate Editor Alex York talks with Janet Ho, the COO of HumeAI. They discuss Hume AI, an AI company focused on developing empathic AI and speech language models for various applications, particularly voice interfaces. Janet shares her background, which surprisingly wasn't initially tech-focused, and her journey from product management and venture capital to co-founding Hume.
Ho talks about passion as a key part of entrepreneurship, the value of hiring people who are not only skilled but also great communicators and the ethical implications of artificial intelligence.
Subscribe to FORBES: https://www.youtube.com/user/Forbes?sub_confirmation=1
0:00 Introduction
:49 Janet Ho's Inspiration For Pursuing A Tech Career
4:58 How Janet Ho Went From Investment Firm To Start Up
17:46 The Latest On Data Manipulation And AI
23:52 Getting Funded As An AI Company
Fuel your success with Forbes. Gain unlimited access to premium journalism, including breaking news, groundbreaking in-depth reported stories, daily digests and more. Plus, members get a front-row seat at members-only events with leading thinkers and doers, access to premium video that can help you get ahead, an ad-light experience, early access to select products including NFT drops and more:
https://account.forbes.com/membership/?utm_source=youtube&utm_medium=display&utm_campaign=growth_non-sub_paid_subscribe_ytdescript
Stay Connected
Forbes newsletters: https://newsletters.editorial.forbes.com
Forbes on Facebook: http://fb.com/forbes
Forbes Video on Twitter: http://www.twitter.com/forbes
Forbes Video on Instagram: http://instagram.com/forbes
More From Forbes: http://forbes.com
Forbes covers the intersection of entrepreneurship, wealth, technology, business and lifestyle with a focus on people and success.
Ho talks about passion as a key part of entrepreneurship, the value of hiring people who are not only skilled but also great communicators and the ethical implications of artificial intelligence.
Subscribe to FORBES: https://www.youtube.com/user/Forbes?sub_confirmation=1
0:00 Introduction
:49 Janet Ho's Inspiration For Pursuing A Tech Career
4:58 How Janet Ho Went From Investment Firm To Start Up
17:46 The Latest On Data Manipulation And AI
23:52 Getting Funded As An AI Company
Fuel your success with Forbes. Gain unlimited access to premium journalism, including breaking news, groundbreaking in-depth reported stories, daily digests and more. Plus, members get a front-row seat at members-only events with leading thinkers and doers, access to premium video that can help you get ahead, an ad-light experience, early access to select products including NFT drops and more:
https://account.forbes.com/membership/?utm_source=youtube&utm_medium=display&utm_campaign=growth_non-sub_paid_subscribe_ytdescript
Stay Connected
Forbes newsletters: https://newsletters.editorial.forbes.com
Forbes on Facebook: http://fb.com/forbes
Forbes Video on Twitter: http://www.twitter.com/forbes
Forbes Video on Instagram: http://instagram.com/forbes
More From Forbes: http://forbes.com
Forbes covers the intersection of entrepreneurship, wealth, technology, business and lifestyle with a focus on people and success.
Category
🤖
TechTranscript
00:00Hi, everyone.
00:04It's Alex York.
00:05I'm an editor at Forbes, and I'm here with Janet Ho, the COO of Hume AI.
00:09Thank you so much for joining me today.
00:11Thank you so much for having me.
00:12Can you give me a 30-second rundown of who you are and what Hume is?
00:16Yeah.
00:17I'm Janet.
00:18I'm COO of Hume.
00:19We are a New York-based AI company, and we train foundation models that generate voice
00:25and language together, also known as speech-language models.
00:29Our ultimate goal is to optimize AI for well-being.
00:33We are building what we like to call empathic AI, and this model can be used and deployed
00:39in any application for a voice interface and generates the right tone, rhythm, and timbre
00:44appropriate to the language and the context and expressions of the user it's speaking
00:49with.
00:50Very cool.
00:51When you were growing up, what were you into?
00:53Was it things like tech and, you know, I mean, AI wasn't really so much talked about then,
00:58but what were your interests then?
00:59I get this question a lot.
01:01People ask me, like, oh, do you study computer science?
01:02Were you into tech all this time?
01:04The answer is resoundingly no.
01:07Growing up, I was really creative.
01:09I loved sewing things, making things with my hands.
01:12I still do.
01:14My sister is very artistic.
01:15My parents are both very creative, even though they don't work in creative industries.
01:20And yeah, I honestly kind of fell into tech, and I kind of see tech as a big horizontal
01:26layer across a lot of jobs and a lot of industries nowadays, and certainly becoming so, especially
01:32with AI.
01:33And we can talk more about that.
01:35But yeah, so growing up, definitely not.
01:38Really loved making things with my hands.
01:41And when I went to college, actually, I studied political science.
01:45I got a minor in computer science, but I also did not focus on tech immediately.
01:49Yeah.
01:50Why the political science route?
01:52What made you decide to study that?
01:55Yeah.
01:56I've always been interested in public policy.
01:57And I do think that it's the primary way we can affect change, especially with more recent
02:02things happening and seeing – I mean, people might disagree with more recent things happening
02:06and how much influence a lot of tech leaders are having on Capitol Hill and whether or
02:11not that's good or bad for the world to be seen, I guess.
02:15But yeah, I think public policy has always meant a lot to me, and politics has always
02:19been a big interest of mine.
02:21Yeah, so I think that's why I focused on it.
02:25Did you see yourself becoming a founder?
02:27Like, when you were growing up, you were interested in creative things.
02:29I mean, there are so many – I feel like a lot of founders are so creative in nature.
02:33Did you want to go down the entrepreneurship route?
02:35Honestly, no.
02:38I also think that maybe that's probably a good thing.
02:41I think that – and obviously, founders come in all different shapes and sizes, but – and
02:48I might get some hate for this, but I actually don't know if it's a really great idea
02:53to have somebody, like, found a company simply because they, like, want to start something.
02:58I've – like, the best founders I've met have always been people who, you know, see
03:03a problem they really want to solve, feel really strongly about that, and are like,
03:07I wouldn't really run a company besides doing something like this.
03:10Obviously, there have been a lot of success stories that suggest otherwise, but that's
03:15kind of my personal philosophy.
03:17That being said, not to say that, you know, one way is better than another, but I did
03:22not see myself as an entrepreneur.
03:24Yeah, I guess I never was afraid to do things a little bit differently from others, not
03:29to sound super cliche, but definitely have always marched to the beat of my own drum,
03:35and I have a lot of ideas, but I don't know if I'd say that I could – I envisioned
03:40myself as, like, a tech founder or, like, you know, leading a tech company or anything
03:45like that.
03:46Yes.
03:47How would you say – because I know you came from a PM background, and then you were at
03:50a VC fund for a bit.
03:52Can you walk me through that career journey?
03:54How have those different roles led one to the other?
03:57Totally.
03:58Totally.
03:59The biggest thing that kind of shook a lot of that up was COVID.
04:03Right at the offset of COVID, I was working as a product manager, and then because COVID
04:09happened, I ended up working remotely.
04:11Everyone started working remotely.
04:13I ended up living in Hawaii for a little over a year.
04:17Oh, wow.
04:18And my friend was starting a VC firm there, and he asked if I wanted to join.
04:22It just so happened that some of the people involved were actually in Hawaii at that time,
04:28and we went, and we had no idea – like, nobody had any idea, like, how long COVID
04:32was going to go for.
04:33Right.
04:34So we ended up just staying an entire year, and I think that actually was very foundational
04:39in kind of a lot of the working relationships I ended up having, Hume and how Hume started.
04:46And so it did kind of happen – it will definitely happen organically, and I think
04:52at every point of that journey, I was just, like, pretty open-minded and said yes.
04:58What gave you the confidence to make that jump?
05:00Like, for instance, from working with an investment firm to then jumping into a startup?
05:05Totally.
05:06So I was actually helping with a startup quite a bit, and then it just happened to be that
05:11the startup needed more and more help, more and more work, and I just really enjoyed doing
05:15the work.
05:17And I was like, gosh, like, I really love these people.
05:18I really enjoy working deeply.
05:19Not that I didn't enjoy my investing job.
05:21It was really interesting, really fun, but I was like, this is something I really like
05:24doing, and I'm quite good at it.
05:26And so I think there was obviously a formal conversation where I was like, well, it looks
05:29like I'm working, like, 80% full-time on this thing.
05:33It was time to make that formal transition, and so I think, like, the time I had to gradually
05:39transition gave me the confidence to fully transition.
05:43The best advice I have for people is to, if you see an opportunity where a lot of smart
05:50people are working and they need a hand, just jump in and do whatever it takes.
05:55I think, you know, I'm not an AI researcher.
05:59I'm definitely not an engineer as well, but I think when Hume first started and there
06:06was a lot of exciting work on the research side, on the technical side within the company,
06:11what they really needed was somebody who would kind of figure everything else out and
06:15keep things together.
06:17And while that wasn't something that I had certainly trained for or, you know, I mean,
06:22I enjoy doing it.
06:23It's about attitude, and it's about jumping in and saying yes to things like that, right?
06:27Like, if you really want to be a part of an exciting opportunity, whatever it is, maybe
06:32it's a company you want to work for.
06:34Maybe it's a group of people you want to work with.
06:38I think figuring out the details ahead of time can often make you feel pretty stuck
06:43and may actually let the opportunity pass by.
06:45I think the most successful people I've met in life are also the ones who ironically kind
06:49of create those opportunities for themselves, right?
06:51Because typically it may be a job that people might pass over.
06:54People might think like, oh, I don't want to do that.
06:56But I think being like, yes, I'm going to do that, and I am not going to complain.
07:00I'm going to really work hard is really important.
07:02And I think it's something that people should keep in mind.
07:07I think that's like the best advice I have to give is just to work hard and, yeah, like
07:14take things as they come.
07:15And I think that ultimately, if you have different goals and different ambitions, those opportunities
07:23will again present themselves.
07:25And you need to be able to kind of prove yourself at every step of the way that you're willing
07:29to kind of take on an existing group of work to kind of get more.
07:33And so I guess it's important to not let yourself be the blocker to those.
07:40Yeah.
07:41There's so much changing in AI right now.
07:43For people who don't know what humans are, or have not used it before, can you give us
07:46a rundown of like exactly who the user is and what they're coming to the platform to
07:50do?
07:51Sure.
07:52Sure.
07:53So we train speech language models.
07:54It's a model that can generate AI voice and generate the right tone, rhythm, and timbre
08:00depending on the context of the conversations of the language.
08:03It's a model that understands speech and language, which is...
08:05We're the only startup with a speech language model in production actually.
08:10And it can also take into account the user's expressions because we actually have this
08:14way to measure expressions in the user's voice and their language that they're talking.
08:18And so what you get is basically a conversational AI.
08:21It's a model that people can access via API and deploy into any kind of product application.
08:26So our group of users are developers who work in healthcare, ed tech, customer service,
08:32et cetera.
08:33And we're actually releasing more tools soon.
08:35But the vision for the company has always been to be the AI voice that people can trust.
08:44And trust not only for themselves, because we have a consumer app, but also that developers
08:49can trust to use in their applications.
08:53And our model can actually call other models as well to do specific tasks.
08:57So let's just say that maybe you want someone to...
09:00You're building an ed tech app and you want the model to be able to answer this user's
09:06inquiry and their inquiry is very technically sophisticated.
09:11It's like a very complicated math problem.
09:14Our model can actually orchestrate another model.
09:16Maybe it's the best model for solving math problems to actually generate that response
09:21and pass it through to the user.
09:23But our model kind of plays that high EQ orchestrator.
09:31And you obviously want this answer delivered in a approachable and understandable way.
09:37It should be delivered in the right tone of voice.
09:40It should be delivered in a way that humans have conversations.
09:44So are your users, you mentioned they have like a consumer platform that consumers can
09:49use.
09:50Yeah, so they're developers.
09:51Is that who is paying you for the service or who else are you kind of working with to
09:55make sure this?
09:56Exactly.
09:57So our primary customers are developers.
09:59They're independent developers, developers at startups, developers at mid-sized companies,
10:02developers at enterprises.
10:05We have a lot of users on our platform and they represent folks who are building and
10:13are technical and they use our model and they can basically add a few lines of code to their
10:19application and be able to call our model.
10:22Okay, gotcha.
10:23So it's in a sense like integrated in the things that they are building and solving?
10:26Exactly, exactly, exactly.
10:27Why did you guys want to go the route of the speech as opposed to just a text coming back
10:34or something like that?
10:35I feel like there's been a lot of conversations about how much should an AI model replicate
10:39a human and a lot of controversy there.
10:41Why did you guys go that direction?
10:43Yeah, speech is just a lot more natural of a way to communicate and ingest information
10:48for humans.
10:50We can listen a lot faster than we read and obviously having visual cues such as text
10:56and or images can be helpful as well, but at the end of the day we communicate with
11:00each other through voice for a reason.
11:01A lot of context is also missing without voice and accessibility-wise it's also pretty important.
11:07I think that's a big reason as to why voice is a really important part of communication
11:15and is an expectation for just how AI of the future will become more helpful to us.
11:24There is a limit to how helpful AI can be if it really is text-based only.
11:27You can imagine in the future we were talking to a company who's working on a companion
11:31robot and looking for a voice.
11:34You might have robotic assistants all around your house helping you with things.
11:38You don't actually want to go over and read the screen that tells you what maybe the eggs
11:43are ready or something's happening.
11:46You want it to be able to talk to you and with that there's a lot of nuance to model out.
11:51We have the concept of back-channeling.
11:53You don't want it to interrupt you when you stop speaking, maybe you're thinking.
11:56You don't want it to interrupt you too soon or too late.
12:02The more that we work on this, the more it's interesting to just begin to realize how smart
12:07humans are and how capable we are of reading social cues and body gestures and each other's
12:14tone of voice.
12:15So delivering that experience is really important.
12:18I think whether or not we build AI in a way that adapts and conforms to human communication
12:27is separate from whether or not AI is AGI, whether it's conscious or has a conscience.
12:36It's really separate from that.
12:38I think at Hume we really see – we really believe in a few things.
12:44First that AI should understand human expressions to serve human well-being and be able to deliver
12:49the right kinds of information to people and ultimately make them happy.
12:53And the only way to do that is really to understand really intent and the tone of voice of people
12:58and deliver that in the right way.
13:00You kind of need that information.
13:02And I think the second piece is that it's very different to kind of have that inevitable
13:10future and also say that that AI needs to be treated like a human or needs to sound
13:16exactly like a human or needs to have a human body and needs to have a human face.
13:21I kind of take a lot of inspiration from the Japanese, I guess, concept of yokai, which
13:28are like these spirits that reside in everyday objects and they're kind of like little demons
13:33that reside in vegetables and nature and water.
13:35I think AI can totally be thought of as such.
13:38They are just different things that exist.
13:41They can be as intelligent – they are more intelligent than humans in many ways already,
13:45but they can be just these super intelligent pets that we have that should communicate
13:50in ways we wish our pets could communicate with us.
13:52Or maybe not.
13:53Maybe you don't want to actually understand what your dog is saying because he or she
13:56may not be that smart.
14:00So I think there's definitely two separate bodies of thought there and they can definitely
14:06be teased apart, if that makes sense.
14:08In terms of the future that you were just talking about, having something notify you
14:12that the eggs are done cooking or that something is happening inside the house, in many ways
14:17it would be so convenient.
14:18Is that something that you guys are working toward?
14:20Yes, exactly.
14:22So we're building basically the voice and the EQ component and IQ component of that.
14:28So while we're not working on robotics specifically, we do have customers in those spaces and the
14:37model that we provide is going to be the intelligence layer behind that.
14:43You want it to have the right way of communicating with you but also be able to say the right
14:47things and just generally be something you enjoy speaking with.
14:52Also can personalize.
14:53Definitely.
14:54What have been some of the biggest challenges in approaching this very new wave of technology?
14:59So many people are after things like this, right?
15:01Yeah, yeah, definitely.
15:02Making AI more and more intelligent and more accessible in your daily life.
15:05What are some of the challenges though?
15:06Because I'm sure many come with that.
15:08Yeah, absolutely.
15:09I think the biggest challenge is how fast space moves.
15:12It's a challenge and a blessing because I actually feel like this race in AI that we
15:16see is really what's moving all of these companies at a breakneck pace to deliver cutting
15:21edge models week by week, honestly, and even day by day.
15:24If you're on AI Twitter, you'll just see new announcements every single day by different
15:28companies.
15:29So I think that would be like the biggest challenge but also probably the biggest boon
15:34for the space in general.
15:36We can look forward to a lot of these applications in production sooner than people expect.
15:45I actually think that we'll be able to pass the AI speech Turing test soon.
15:49You won't be able to identify whether or not an AI sounds like a human or is actually AI.
15:59Do you think that's a good thing?
16:00That's a great question.
16:01So at Hume, we actually are pretty ahead of the curve on this.
16:04When we started the company, we started a sister nonprofit called the Hume Initiative
16:08and it sets forth the only concrete guidelines for uses around empathic AI, so AI that understands
16:12human expressions.
16:13The two things that we don't allow with our technology are manipulation and deception.
16:21I think they're actually really great rules of thumb because almost every single bad use
16:27case with AI can fall into those two categories.
16:29We talk a lot about the future where people will inevitably have AI girlfriends and AI
16:34boyfriends.
16:35Is that a good thing?
16:36Is that a bad thing?
16:37Let's look at this framework.
16:40If people have an AI girlfriend or boyfriend, are they being manipulated?
16:44Are they being deceived?
16:48How do we define those things or are they entering into these relationships very willingly?
16:53From a moral philosophical perspective, where does the onus of the company come in to say
17:01that, no, you shouldn't have these, we think, unhealthy habits versus you are a consenting
17:07adult.
17:08You know exactly what you're doing and you're not being manipulated or deceived.
17:12I actually think that while there's a fine line, there is a clear line between applications
17:17who have good intent, want to better the lives of users and improve their well-being.
17:25That can look like many forms.
17:26Then there are also use cases where the intent is to manipulate or deceive the end user with
17:31AI.
17:33Those are the ones that are actually very, very problematic and that all AI companies
17:39should be working towards ensuring don't happen with their technology.
17:42I feel like a lot of that has come up, especially with the past election cycle.
17:45There have been a lot of cases of manipulation, like you said, of either data or putting it
17:50out in a way that maybe someone didn't consent to or whatnot.
17:53How can users, like you mentioned, identify if a company, if a platform is doing it for
18:00the right reasons or if they are being manipulated and deceived?
18:02Do you have a set of tips for users out there today that are just kind of coming to terms
18:06with using AI?
18:07That's a great question.
18:08We talk about this a lot at Hume.
18:09I think the best way to think about it is, is what I'm using improving my well-being
18:13in the long-term?
18:14If you're being deceived, if you're being manipulated by AI or maybe you don't know
18:20it's AI, how is it affecting your day-to-day life?
18:24Maybe you're being scammed and you're losing a lot of money.
18:26That doesn't sound like it's making you happy in the long-term.
18:29Or maybe you're having this relationship with an AI persona whereby in the short-term you
18:34feel happiness, but in the long-term it takes a toll on your mental health.
18:40That by definition is also not aligned with your well-being, it's not serving your well-being.
18:47So I think understanding well-being as a holistic way of looking at your life and seeing whether
18:54or not the tool you're using is helping that as opposed to harming is probably the best
19:01way we can do this.
19:03I think for AI companies as well, they should also be using this metric, well-being metric,
19:08not even immediate happiness.
19:09There's a lot of things that give us immediate happiness, right?
19:11But it's like, especially with TikTok, which I'm an avid TikTok user, it makes you immediately
19:17happy to be on it.
19:18And then like three hours later I'm like, okay, I got to put this away.
19:21Exactly, exactly.
19:24So I think that's very important.
19:25And I think especially for adults and just older users, that's something that they're
19:32able to really reflect on and decide on for themselves.
19:36And they have their free will to do that.
19:41But I think especially for younger users, that's where it gets really murky, where their
19:45brains aren't fully formed, they probably don't even understand or have the ability
19:50to reflect internally, like, is this making me happy in the long term?
19:55And I think that's where AI companies actually have a lot more onus, and parents do as well.
19:58But also, it's definitely a very new and prickly situation.
20:03And what about government or policy in general, coming from a background of being interested
20:07in these sorts of things?
20:08What are your thoughts on regulation from the top down in that way?
20:11That's a great question.
20:13Because at the end of the day, I do believe in, you know, I think like, policymakers have
20:20a lot to consider.
20:22I think regulation has, you know, a lot of waterfall effects, a lot of rippling effects.
20:31I don't think it would make sense to put onerous policies on AI companies for national security
20:37reasons as well.
20:38As you can see, even though the US is at the forefront of the AI race, that may not be
20:44the case very much soon.
20:45And I think it's very, very important from a national security perspective that we maintain
20:49our edge, and the free market economy, and everything that the US offers for the brightest
20:56minds around the world is the reason why we are even in the position that we're in.
21:02And so I think regulations need to be very, very careful.
21:05We don't want what's happening.
21:08Yeah, we don't want what's happening in other countries or other continents to, you know,
21:13happen in the US.
21:14So, and I also don't believe that our policymakers, unfortunately, have the right context or knowledge
21:22about these nuances.
21:23I think that our government has historically been pretty tech illiterate.
21:27And I'm sure that, you know, as consumers, we've kind of seen that play out in a lot
21:30of Senate hearings.
21:31And just generally how our policymakers are not very much, they don't really know what's
21:38happening, you know, in tech.
21:39And so I think it's doubly dangerous that those people would even think about making
21:44policies that maybe restrict, you know, companies to operate in certain ways.
21:50That being said, I do think that it's important that, you know, the government gains that
21:57technical literacy, especially with the pace in which AI is developing.
22:02And I think that there is no other way but to work with the leading minds in this space
22:08and business leaders in this space to do that.
22:11Obviously, everyone has their own interests, so I can't really, you know, help with that.
22:16But I do think it's really important that the government gains that technical literacy
22:19first, understands how policies can impact everybody, not just the end user, before they
22:28kind of put things into a blanket effect.
22:30Because even if they were to put a policy out there, if it wasn't really well thought
22:33out, it's not going to be very well enforceable either.
22:36And we've kind of seen that in the EU as well.
22:38Yeah, definitely.
22:39I want to talk about this like from an investor perspective as well, in terms of like understanding
22:44these new technologies that are coming through and really supporting you guys in this.
22:50You guys have raised $50 million.
22:52Is that correct?
22:53Yes.
22:54What have those conversations been like?
22:55How eager have investors been to put their money in, and do they have any hesitancy about
22:58like what the future of AI looks like?
23:00Totally.
23:01Both of our rounds, our Series A and Series B have been preempted, so we actually weren't
23:04actively fundraising.
23:05Investors have approached us and offered very, you know, favorable terms for the company.
23:12Investors are very excited about and are very intelligent and smart to the trends, right?
23:21They know what's coming.
23:23Their goal is to imagine the future and fund the ideas they think are going to be very
23:28successful products that will improve the lives of everybody in the world, hopefully.
23:35And or, you know, not just improve lives, but also solve real problems, which will obviously
23:40improve lives.
23:41So, yeah, our investors are very smart individuals and they're very supportive.
23:45I think that they really believe in the strength of the team and the vision and the problem
23:49area.
23:50And yeah, so they've been great.
23:53What is the money that you guys have raised going toward?
23:55I know building an AI company is just expensive, like data and research and value-wise.
23:59Exactly.
24:00You're exactly right.
24:01Yeah.
24:02So I think most AI companies, especially AI companies that train their own foundation
24:05models will let you know that most of their money does go towards, you know, research,
24:10whether it's data and or compute and are training the models.
24:15So yeah, definitely.
24:16That's that's definitely it.
24:17There's no nothing special there.
24:19It does go towards things like that.
24:20Yeah, definitely.
24:21What are your visions for the future of Hume, of the AI industry in general?
24:26What are you guys preparing for right now?
24:28Yeah, absolutely.
24:29Our vision is to be the de facto voice AI for all applications, all applications that
24:37developers are building, applications you have on your phone as a consumer.
24:41It'll be a voice that, and we already have a model out, EV Empathic Voice Interface 2
24:46that does do that.
24:47EV3 will be launching very shortly in the upcoming weeks.
24:52But you know, we want people to be able to open up their phones and have something that
24:57they love talking to, that can help them very easily, that understands their preferences,
25:02is catered and tailored towards their well-being and their long-term goals, and can, yeah,
25:08be a friend.
25:09So in this fundraising journey, the AI industry is obviously so saturated.
25:13Every AI company is coming to investors right now asking to be invested in.
25:16What do you think set Hume apart to really start generating this funding?
25:21Yeah, absolutely.
25:23The first is, I think, building the right team to position yourself for a really big
25:27opportunity.
25:28You know, funding is a part of an opportunity formula, right?
25:33Like funding equates to a portion of the company, which equates to the value of the company,
25:38which equates to the value of the space and the market.
25:40And so I think that it's important to have the team ready to, or not the team ready,
25:47but the team and the product and the history and the work to equate to ultimately the market
25:54opportunity and the problem space.
25:56And so I think building that team is very important and making sure that you have the
25:59right people in place.
26:00And then also working on something that's really great as a product that has a really
26:07strong capability to meet that demand and meet that problem.
26:12And then related to that, working on the right problem, working on a problem that's big enough,
26:16right?
26:17Like it doesn't matter if you have a compelling product to solve a very specific use case,
26:23if there isn't just going to be VC interest in that and, or if the problem is too small.
26:30And so, yeah.
26:31How did you guys go about building that team?
26:33What were some characteristics you looked for in people or roles that you had to make
26:36sure were filled in order to get that funding?
26:38Totally, totally.
26:39While I can't take credit for the technical hires at the company, which are fantastic,
26:44I have been a part of many interviews.
26:47The team is amazing.
26:48And if people are interested in working at Hume and you're an AI engineer, please apply.
26:53But hiring is its own different ballgame.
26:59I like to think we're pretty good at it at Hume.
27:01And I think that's one of our strengths as a company is that we are really close as a
27:07company.
27:08Everyone really enjoys hanging out with each other.
27:09We spend 10, 12 hours every day in the office together in New York City.
27:13And I think when you're hiring, you want to make sure that they're a really great communicator,
27:20they have the right skills to obviously fit the criteria, more importantly, and then that
27:25you want to work with them.
27:28Hiring is super interesting.
27:29And I think that the more that you hire, the better that you get at it.
27:32It really is people reading and, yeah.
27:35Yeah.
27:36And my last question for you is being part of the under 30 list, being a young leader
27:40in this space, what do you think youth has to do with?
27:43The growth of AI right now, how do you see your position as someone young, but building
27:47in this innovative industry?
27:48Oh my gosh.
27:49I do think that the younger generation is so critical to this.
27:56I've highlighted one of the concerns, which is that because the technology is so new,
28:01and when we talk about youth, we're even talking about maybe folks in middle school or high
28:05school.
28:06They are a lot more tech savvy and a lot more quick to adopt new trends and new ideas than
28:11older generations.
28:13And oftentimes, these tools just don't have the safety guardrails built in yet.
28:16And so that's very dangerous.
28:18But at the same time, their adoption, their willingness to test, and their ability to
28:25utilize these technologies in their lives is really what's accelerating the market.
28:30I think that since I graduated college, chat GPT and open AI came into play.
28:35And I can't even begin to imagine what school is like now.
28:38I've heard it's quite crazy.
28:40I just think they don't have tests anymore or something.
28:42Something crazy.
28:43Something actually insane.
28:45I think my younger cousin was telling me, he's like, yeah, we just don't have homework
28:49anymore.
28:50I'm like, what?
28:51That's crazy.
28:52And it's like, well, yeah, because homework can all be answered by AI.
28:55And so I think that we'll definitely see more and more assistive help with AI, things that
29:04will make everyone's lives easier.
29:07And I think it'll be really important to build the guardrails for younger folks.
29:12And also, if painted very positively, which I believe I'm like an AI optimist, I do think
29:18that the work, what children and what younger people will need to learn is probably going
29:24to be quite different.
29:25Like they might just need to adapt to working with an AI and be like, okay, well, maybe
29:30I don't need to know how to spell things correctly anymore, but what are the things that I can
29:36build with AI?
29:37Right?
29:38And maybe it will allow me to build my own website with AI because it's super easy.
29:43Maybe it can accelerate me starting a business because it lets me find the information.
29:46And so like the things that you will have to do in the world still don't change.
29:52Still need to get a job, still need to start businesses, like not you, but like people
29:55in general, like the economy still has to run, but I think that it'll just be more assistive
30:02to those individuals and to those, the younger generation.
30:05And I think for schools, they need to figure out how to adapt their curriculum and let
30:13people work with AI to achieve even maybe higher expectations and higher performances.
30:18But that's not easy because it's all happened so fast over the past two to three years,
30:26but it's exciting.
30:27Yeah, it is exciting.
30:28We have a long road ahead with AI, but so many cool things happening.
30:30We do.
30:31We do.
30:32Well, thank you so much, Janet, for joining me today.
30:33It was awesome to hear your thoughts on everything and how you guys are building today.
30:36Of course.
30:37Thank you so much for having me.
30:38It was so great chatting with you.