Jing Conan Wang [07:57]

I am very excited to have Matthew joining us today to share with his reasons. You have incredible experience, I have to say. So. Man Show is currently a founder and Managing Partner of Machina venture, and before that, Machina Ventures is in investing in 19 AI that amplify human attention. Before that Machina video, please answer the strategy at Google in which he invests a lot of the mission driven projects and with high social networks, a lot of cool products. And even more, Machina is like a nature in the burning well, you feel like you have a lot of time, and even more, you have a JD from UCLA. And even more, you also engineering, of course, yeah, yeah, cool, cool. So that guy today, so we actually like a lot of questions before and thank you for so a lot of our attendees, they are very interested in your early journey and career Foundation. And I heard that you are a game developer when, when you are young, and yeah, chat down small about that experience.

Matthew Stepka [09:48]

First of all, food story. Hello, hi everybody. But anyway, here, and hopefully we have a good conversation. I was trying to be as vulnerable and open as this is great forum founders. I've been a founder as well. I've been a work a lot of founders. And it isn't a really hard while. It basically has to be insane first. I'll start with so you're a little bit crazy. Works wonderful as well, but it's a very hard road. I'll share some of my thoughts and experiences about that. Yeah, so early on, when I was 14 years old in grade school, my dad was an aerospace engineer at NASA. He wanted to be a cartoonist. That was my dream. Wanted to do cartoon. My dad gave me a computer, a Tory 400 computer with they called it a McDonald's keyboard, keys, McDonald's entry terminal. But anyway, so I had this computer, and I learned a program I basically didn't have any comments that went on. But I love that computer. I knew it. I started making video games on Atari and to the point where job offer at 17 based on a computer game I was just still out there, I gave away for free called sharp. Recently did a review of this through the AI. Anyway, so I did go to college instead. I decided not to take the approach to play before this call back and live here and also a business career as well. But interestingly to me, video games at that time, this is around working on movie workings, and also Sean and the idea that this other world existed, you could create, and your other peers could create, or imagine. You could deal with computers to manifest it that way, you felt like all God and your young person, that is amazing. So yeah, I think that's the high priority.

Jing Conan Wang [12:19]

Yeah, exactly, yeah, yeah, and so and like, you sing it out of box, do something. And I also heard that you didn't go to the traditional engineering class. You end up being a Jing for

Matthew Stepka [12:44]

Yeah, after doing a Jing period, to college, and then I decided to California, Ohio, and wanted to go to school first. AI program. We go to private school, and it goes well over business school. Actually, I think it's always good to follow your excited about rather than when we think to follow you're excited about it, but what you think you should be doing, and I know it's hard to do that, sometimes you say, this is what it should be doing, and the philosophy of philosophy, But how do you take the values of society? I society and make a set of rules, kind of program rules we follow that a very imperfect expression of those values. And write a lot, and there's a lot of states practical system, but you learn that the processor would, in some sense, a very mobile profession. I know a lot of people don't like lawyers, but she has really inspiring. Never back the plan.

Jing Conan Wang [14:16]

Yeah, yeah, exactly, exactly, yeah, yeah, and also, as you mentioned, about the animal practice in law, but also, I heard that you found the first internet company, is that something is that project?

Matthew Stepka [14:33]

Yes, when I hear anybody Law School and the Quincy still in between there, and we're all for big years, the internet starts happening in 1995 and little cafes. We love the internet team. Online was one of the nodes of the internet. I was just there two days ago, and we thought cool. We started an internet cafe, which was the stupidest thing we could have done anything else. It was the first on the West Coast. There was one in New York, in London, started Matthew park here, um, that's

Jing Conan Wang [15:29]

pretty cool. Um, by the way, I call they found a cold quality Illinois, like coffee house. It's also a copy. So,

Matthew Stepka [15:38]

yeah, so many revolutions and new ideas and pulling it in hop houses and bar so this is coming together physically. I don't know how people never be replaced by it seems that that's when people come together and get them to make new decisions and bounce off these ideas on each other. Be bold and bold with each other. Do something together, start a new revolution,

Jing Conan Wang [16:10]

yeah? And you like creatively combine it now with cafe, yeah? So that's also backing to like the native who joined Google. You have incredible career at Google, and here's the that moment, what happened to your life people around them, what made you decide to think about

Matthew Stepka [16:35]

joining Google? Had startups, and it was a little bit like how it's doing happening AI now, easy way flying around. It was kind of hard to know what to do with the values. I did go with E commerce for travel, and it was commerce decision Stanford, 20 people at the time, so it's hard. More. Realized environment was valuable to make sense. I did feel like travel was doing one category, and the company, we did pretty well, and the CEO of the company and leading it through a process of the we talk about us announced now doing a situation when it's not like this, it's like this where you're just trying to buy, just trying to make people really interesting law, like the CEO or CFO. If you don't make payroll tax, you are personally, we're going to make cable tax every time we have a tax. So they're like facing that now to make sure that just everything loses trouble. Just making something here, this is going to learn when you got all this money. Machine learning was great, but it's also at that time is horrible. Beginning, about two thirds of the people that economy and was hard hiring, and they have students who know so Hard people think we hire people, lies and So,

Jing Conan Wang [19:41]

yeah,

Matthew Stepka [19:48]

lots of mistakes, and they said, Google a couple times to do it, but finally, that seems to be a while, But then after that, I came in at the time, downside, but you

Jing Conan Wang [20:27]

never know what is the next journey you will have, right? So it was, I think that point is probably with a little time your in your career, but then you start the new chapter in Google. I think it incredible money at Google, a lot of high impact missionary products, and looks like you also didn't take a conventional job at Google, like, well, most people either the more traditional branches, advertising, commercialization part of the branch of Google, but it shows the difference. What is your thoughts, too? So

Matthew Stepka [21:07]

what happened when I came into Google? Here? Internal strategy group called this operations, and even the director, initially, because I had a very kind of heavy operational background already, I was kind of an all kind of person, and that was helpful they want to start new initiatives. So they were trying to figure out something Africa. They wanted to help remove Africa, return to Africa. That was kind of a multi dimensional problem. We had a lot of challenges, lots of opportunities. And also they want to help Africa. So the whole initiative driven, but also business and so consultants trying to figure it out. I had to go back on that. Then about a year later, founders, they wanted to invest in renewable energy projects do that. So we went for it after a couple billion dollars. That money into a renewable energy budget. We were largest wind investor two years ago, so we were investing projects for wind and solar later. Eric, when I drove to to live in Afghanistan, it was Iraq with the State Department, and he got into Shanghai from skate park at the same started Google Ideas in Google now using technology, hopefully to help people get their Freedom, how to get information. Help also help Congress. Well. Help also shift a lot, especially to video, situation, progress, net, first level, I see open as close to the mouse. So anyway, add up Google and then actually this would be enough. Pragmatic and not always easy to be American and organizations can have lots of baggage issues. You know they aren't gonna work properly, even if you see behind you. So you have to just kind of keep making the progress you can make, and take the list you can take, and always keep moving forward. Use force analogy, but moving forward as much as possible. And then again, that's what's useful from senior management, if you can do that, that's progress. So hard to get to move. It's like cancer. So hopefully you can help turn that field and privacy with attitude people to fire. It's hard because we were working relationship technology. Larry Page was gonna come and talk to the team hostage, and they're working on the digital cities and optimize traffic and resources. Larry used to talk I don't even talk everybody going Mars, how we should decide what the government should be in Mars, We're working on computer between discontinuity between,

Jing Conan Wang [25:21]

yeah, cool, cool. And also, many of us also cares about, like, you have a great done, great, a lot, great article. But what is your biggest pain, anything, and what is the main learning or gateway

Matthew Stepka [25:39]

that I really probably the biggest one kind of route to so we invest in a company called virtual B, which is a Saturday company. We bought Internet access to Africa company, and founder very well and entirely founded Google. He is a checker out so they're hard to do that, and so they hired him to do that. And he building, he has 15 engineers building design, waste, the whole thing, and then a little bit more process. But in the end, he ended up meeting, working with another company, and then not staying in a very long so I think we would do that project. We live in the faster, sooner, and I actually think better stewards of medical health within the world. I love Starlink. It's awesome, which was at Google this opportunity, also for google.org It was always a challenge. We number of owners within the board and owners within the work. And it started from way over high education. People want to Google or do everything. And this word started on founders, because the founders of Google, in their ideal letter, said, We want to Google that word they call the Google location at times, to eclipse the Google itself and impact, which is kind of impossible with a team about and so this is a way to do that, extremely cool, and also how that was challenging, making a much more modest school to make it a very Well one nonprofit, and we separate out the engineering team, and that became part of Google then And last, we moved around it more appropriately. It felt a little bit like we defined succession too. Success.

Jing Conan Wang [28:09]

Yeah, cool, cool, yeah, and news time you have a good time to Google, but I had a long learning but I ended up also chose to go New Jersey and became a VC. And what prompt do you do pursue a new journey to become a VC? Yeah, sometimes I feel like

Matthew Stepka [28:32]

things just happen. You want to hear some race vision, honestly. So it moved on me, personally, I host a lot of events my house. Like, I love bringing people together. I've always done it, even though. And then when we get together and things just happen, people meet each other, and so your company, or maybe a motivation or not. Often, whatever you do something and so, and when we're sitting around fire again, there's no say. So before long as time, I stopped doing it more seriously, and I started as an angel and being more of a professional investor. Now I more of a professional investor. Now, anyways, I might as well do that seriously. The hardest part is saying no all the time. You say no most of the time, especially 100 people order at least, at least now, and I think that's I mean one of the moment I realized what it means to Get out there with an idea just about there. Do you think you know the idea is keeping up at night and you have to stick it happen? And he had a very pushy job at the base of the guy and asked, financially. But if you really can't not do it, then do it. It's fine, just realize, but just fight it anyways. And there's sort of course, but to start stressful, emotional roller coaster, and depends where you are the person, there's some parts that you're not gonna like, no like operations part, I need a clo for sure. Like, do a weekly immediately. Yeah,

Jing Conan Wang [31:05]

yeah, yeah. Do you still remember the first game like the most memorable

Matthew Stepka [31:14]

one? Yeah, beginning as an angel, I was mostly just talking to my friends and did a lot of the best. With my friends, and a lot of investing in friends, which was, I got a couple lucky ones, but learned to say, no, it's really hard.

Jing Conan Wang [31:55]

Yeah, yeah, yeah. So also, like, curious that have you deal with like difficult situations or difficult founders, and what's the most difficult accommodation you have dealing?

Matthew Stepka [32:10]

Yeah, but it's interesting what the founders used for you started playing most cases between companies that did do that, and one person Dealing with a lot of personal issues before, had a very good idea. They couldn't tell immediately at all. They needed to learn basic company versus company, and that is a big need for some people, and so I have the ability to do a counselor, hire more people, maybe narrow the scope, kind of helping them to rationalize the business that they might think that they interference, or Whatever. If you don't hold that trust first, I'm really trying to help you be successful. I would like to think that the person is responsible, and it's hard for the belief to be able to get even money. And, you know, I feel under and, yeah, do they don't have, do we need to do some emotional closure. So that was an interesting process. It gets very personal. I mean, this is people's lives and wines and cases. Lives in terms of their good identity and sense of worth and their dreams.

Jing Conan Wang [34:12]

Yeah, yeah, exactly, yeah, because, like, that's kind of like investing a city perspective a lot of dedicated situations with oil and energy and today, we actually have a lot of founders joining today. And I want to just go preach, hands up if you are a founder or about you see AI probably 76,000% and hands on media currently in the boundaries, are you looking for boundaries? Cool, quite, quite big and a decent question for you, actually, for a lot of boundary scale, we are also like doing boundaries, which is very, very important, and the current environment hasn't been very friendly for the fundraising as well. So what would it be your audience to founders who want to do fundraising or plan to in the near future?

Matthew Stepka [35:14]

Well, the first thing I want to do is, I think especially towards the China protocol. Try to get much full convince yourself as the use of your client the most valuable thing you have in your life, you have opportunity or in your career. How do I start and getting rid of, get rid of many risks at hand in the process. Because once you see or any investor would strong position, one of you will know that you actually talking about and you thought through all these potential risks. How do you plan? You will get questions. You should be careful. You should be offended or annoyed or competitive. You shouldn't answer and also have a conversation with people. Are engaged with them, and they make the points that concern. And here's some ideas and difficult things work. You know, just exuding confidence is not pointed to it. Just write confidence, no matter what elder stuff wrote out there that more applied, you would have thought through very well and defensive and real world care, real world, believe it, the problem is willing to have that, you know, conviction, but be a Very good listener and process. And company, if you get revenue from your customer, company, all the better government grants renewable Energy speed until Tuesday, why you're making money, raising money. Doing, what do you need? And I think that too very carefully, and how do I debut things To the chance to do this? So very solid arts going to find this out by the chance that they will learn something together, even available. It's a good failure. Good failures are great. Bad failures. We didn't think about it in the past. And think about what you're doing, why you're doing money? Don't know why? Long term

Jing Conan Wang [38:52]

answer, yeah, that's really helpful also, and you have very, very interesting systems like the United AI. Can you elaborated more? What assistance behind the commissioner?

Matthew Stepka [39:12]

Ventures have a bit of a impact. Focus, going to make money, but to make money, but invest in AI that we can go and we're controversial. A lot of people that are creative us to do jobs, even the environmental energy and of course, that question so there's a lot of ways to be concerned about how AI applied. I think most important thing is that mind AI is keeping the human center of Asians generally. These two things, one, you'll help to have a better experience for the person who's involved, and they won't resist the use of AI. And then a good example is when they invest in made an AI, a co op or a COVID for doctors in the hospital center, and by calling that, apparently it was just held to all this documentation, also make sure you don't miss the same protocol to make sure they don't follow it. So even mind, AI makes things so that people feel more more human, to have more time doing the humans to watch the doctor can hold your hand and talk to you. You can see how you're doing the way that AI does a show and using their intuition and also the care. We don't know why it does, but we know what it does. The placebo is an amazing thing. We don't know why it works, but we know it does. We also the doctor told you, we're going to be okay. You're more looking going to be okay. Eat a little bit. So those kind of things are important in general data. So lied to your patients. But to your picture, the human center not only yield to the results, but it also makes it really hard to dislodge. So it is the most automatically. So all the build decisions, all the design decisions, your workflow, so it's really easy and comfortable. Gamify, it, keep in person, engage all those small learnings are very hard to see or implement or not for a while. So we feel deep with these organizations and these workflows. Inevitably. That's fine. So things always work that way. Sometimes good thing in the world. Because,

Jing Conan Wang [41:46]

yeah, yeah. So that's actually very interesting. That's a lot of notion about AI and like, it's also the urban hot topics. I'm curious AI, what is really excited about AI? What do you see about

Matthew Stepka [42:06]

fusion? AI was in college, actually, I took a class on AI, actually, classes. On AI, and it was cool. Areas was mostly expert systems, machine learning, think is just a pipe. The potential is there, and I think for a while, and Google people lives with the the project group today, Google lives at the vaccine, and we did potential there. Google has the most had, and it has the most main people in AI, but they weren't able to make homework that philosophy, and

Jing Conan Wang [43:26]

so what do you see like the biggest challenge ahead by being the whole evolution of technology, on what you see on the industry?

Matthew Stepka [43:37]

I do think the near term risk is that AI we use by the bad people to do that. I'm actually surprised it did not lay out in the industry. That is surprising. It is actually a radio factor to the battlefield in a way that I find disturbing. That's one area Austin I expert us not to to build AI with people to good. Let's say our security. HTML, AI is going to be a bad little box on both sides. Ai with each other. No guarantees like the dependent history. So I think we should work ourselves so other things worry about using AI intentionally purposes is for most people not having literacy, making AI magic. You know, it understands this, or it's aggressive, or it gets needed, we're kind of easy to fool that we should be very cool. It's really easier to look for humans. And in the end, I think biggest fear media people trust AI too early, and they should, as far as it understands. And this is a good example of this is today. Judges are using AI to make some decisions like whether you fail, whether you go to jail, five to 10 years. And this is being used today, and these systems are extraordinarily biased, and even more bias, it's a very humane thing. Human being should be involved accurately incisions to put another human being in jail. So I think we should be saying, pure use that tool, but make sure you don't get over here by natural AI five years the judge. But I think those are really dangerous, suddenly subservient. AI is because we're either lazy or we all see things really smart, especially if you're living especially is children and young people. I mean, people don't have emotional intelligence, and it does not experience and they're at a heart rate. Doesn't know what human experience really is. It's just better. And these groups taken laser hearing systems, worried about children being positive AI and having them raise their kids.

Jing Conan Wang [46:56]

Yeah, yeah, that's exactly yeah. So that's a lot of considerations we have. One thing about technology evolution, all the backability in another round technology revolution, what we see, and I'm pretty sure you sharing, very careful and also like luck, even at the audience will certainly open the mic to everyone ask questions, to think about that. But before that, I will have a question to you that imagine that you see you allow time to be back in 20 years, but now you're younger than putting it no and you now bombing goes back. So what exactly, well, games go back

Matthew Stepka [48:00]

to still like a word go back to because it is the core possibility. Also, games can be used to do a lot of things, identification, a federal company that uses games to help treat ADHD. So I think we should do games as drugs. We should do the best way to change our lives, affect our medical state or political state. We know they do. We just don't get prescription for it, but we should really think, but it's a very powerful tool. It can really help people.

Jing Conan Wang [48:41]

Yeah, exactly. I think is actually pretty interesting field. I'm also very intelligent. Cool, before we try to monitor all this, last question, do we have any recommendation about books or resources and share with the audience a

Matthew Stepka [49:00]

book I read recently, actually two books. They're actually not recent books. One is called GED guter, Azure block GED. So the book I think I read partially at least about 20 years ago, and it's still a total gradually book, very thick all but it talked about how all these things are interacting together in different fields, and basically talking about how having loops of paradoxes, and then how your brain can actually deal with this. But actually, today, all of us cannot. We can't just circle a lot. We had to kind of have a lot of brain circle. The humans deal at all time. We are very completely creatures in our brains. And that's probably because, and that's them so magic, so magic in that other two posts, maybe one is finding well and great working views on the nature of Success. Really did all the hard work about being in the Philistine. Very interesting concepts.

Jing Conan Wang [50:29]

Awesome, cool, cool. Thanks for joining. Any questions for the audience. Thank

Matthew Stepka [51:06]

for sure, because this early stage is harder to be too hot data you typically look for the space that will be actually appropriate, because it ultimately have reasonable learning algorithms themselves so pick this learning algorithm so people who Are too comfortable sometimes are too comfortable against you sometimes, but all that person and understand their information layer, so there's some, usually not too much technology risk. One company I'm looking at is doing a full stack approaching item, and a way to do it is very novel, and so I brought people, and it's not you might solve the problem we had. So Katie, do it applaud the idea that people are investing in things like Mars or infusion or quantum computing, but it's not kind of magnetic. What about a product? Market? Product is where they are. So if I'm investing pre seed, they may have some example of that. They may have a feedback here. So more of us setting up that test. They talked about and the way to see the big connections with a AI detection. And they did pretty well there. The mentors really good. They moved into other areas. They got aI governance. They pitch as well and added more to the product that is now they have a nice, big enough set of features, not just a one feature. Typically you to do another part which Thank you. Any other questions? Think about

Matthew Stepka [54:02]

they is that moving a handful of companies that will provide lost services open AI, people are kind of popular to come outside, unless we have some paper through. I don't think it wrong with the question, but Google definitely challenged. Now, the whole challenge Google has is the cash machine over here, and they had this when I was there. The cash machine is advertising, which is based on an imperfect search engine advertising. And that was true back then. This is the search versus the best problem. If you know the right answer, if you tell the consumer, then you don't have the problem with that, and no one's paid right answer. So that's the problem. They're probably they can't charge enough right now. I'm like asking ridiculous questions. Make me a letter and chat with two 3000 for the answer, and that's and they're not making money. They can't charge the money basis. They can't do it in a package. So on the application layer, ultimately, is using the AI to do incredible things. That's where the business will be, and I don't think we'll be making money. And day, that's how I see it playing out. Do warmly on computing at a big level for a startup now, in time and ways, will improve this kind of on some word law, kind of wasting your time so you can figure out something that works, and kind of economic, right now, years now, it's interesting. So decided that they are adding enough value, and they have solutions. Now you could have sort of taken advertisement moving into the answers. You can see that brand advertising, kind of YouTube, there's no solution out there. Yeah, us all time, all time. And I think being that close another line to suggest things. AI had an incredible position to be and have trusted word. Yeah, data

Matthew Stepka [58:44]

is it? AGI as objective? I don't know that. Yeah, first, it

Matthew Stepka [59:02]

say, my Transformers will visit there to true HDR. Peter work said recently that we really have ATI, or part of it, people learn more efficiently, and robots too. The baby learns how to walk a million times, defending the robots learn a lot. The simulation is not called a million. Simulations. So we know it's not very efficient. We know that these are not the most efficient way our brains use very little energy, extraordinarily efficient energy. So we really believe comparison of synapses to edges and roads to neurons is a very rough estimate. These are not really long that's good potential, and they're not there yet, and we're probably gonna hit the shoulder of the expert technology now we're now last opportunity to one apply we have today. Just make that work. Doesn't even decades to figure out the value. Secondly, working on making what we have now just a ton more efficient, right size in the problem model to the problem, maybe to do my recipe for dinner, because smaller we my recipe for dinner, that kind of thing.

Matthew Stepka [1:01:20]

that question? Any suggestions, please? These are often insecure themselves. There's no way to answer, so don't take it too seriously. It's been everybody and kind of like a shark tank thing. And classically, everything has first competitive advantage. And then I think, go on a little deeper. Why is it moisture? Why? For example, you example, this is medical adaptation. First of all, is first because there's so many small learning at the table, and there's no way to do that faster or see what works, and you can work with their data so that it plays very well. There is really doing it, and the best data is decades to actually the doctorate or advocate support to grow it has also need to be to see kind of model. So not only that, people want to do more. Hey, should do this too. Before long, we're using this almost completely viral consumer app, so we kind of want to have many more classified data there, rather trying to have one answer to shut it down. Because, first of all, and I think this would be probably an English online. They're never going to be hurt, for sure, question, but they definitely have,

Jing Conan Wang [1:03:16]

yeah, in any stretch of time, maybe I have two more questions. And yeah,

Matthew Stepka [1:04:37]

other PCs is hardware or software.

Matthew Stepka [1:04:53]

manufacturing. So movement. So one thing is, it's hard, so work with Google Analytics team quantum dynamics. So a lot of this software group. So first of all, I like your call the team vertical coach, maybe do like that in general, or one or two things deliver real value, especially if you lead me play some customers on think that's a good approach to take as long as you show, hey, here's one. That one is that strongest pool, whatever reason, at least a good approach is that, you know, I yeah, you couldn't iterate quickly. Hardware is hard. I do say I suggested it for me tonight. I just, it was a couple of, I thought, mostly hardware and even that was fair. When company invested just went through a transformation, just a soccer company, position, and even use that class as a case study software speed, that's great, if you can find ways to do that. If you can be the Android, things like that. Most as possible Harve.

Matthew Stepka [1:08:48]

Yeah, using medical research. Oh.

Matthew Stepka [1:09:22]

I think so medical research is an opportunity medicine and doing data analysis just what I want to leave, because they don't have the medical background as much of that exposure consultants and pharma companies. So not to say that there's lots of opportunities. And I think what's cool about AI is it has be used to see things that humans don't see, some relationships and so forth, and just do the numbers is so overwhelming the data that we can get most farm company basically cooks and go into the lab and just get together and see what happens. It was really great, and only over time became much more systematic. And say, you know, actually, we're trying to go back to this idea behind it, and then now, with AI, much more process driven, which means they can take a lot of money. That level exists Braille, which use AI to find relationships, DNA with tumors, cancer, blood tests. But there's also some videos, and it's a hard road. It's hard to find out so moonshot, I question

Matthew Stepka [1:11:24]

Investors here versus China. Yeah, good sense for the community in China. I can't wait to see how that compares some of the valley theory, introducing more plastic space is more open San Diego program, electric rising is based here with some Stanford professor. So lots, even, I think a certain amount of luster that comes from this area. Yeah, I don't know. I think better word company. Yes.

Jing Conan Wang [1:12:17]

Thank you very much, Marshall for sharing so much vision for us. So in the interest time, this will be the final questions for the fireside chat, but it's not at the end of the day. So after that, we will still have a lot of time to mingle, so please stay as long as you would like, but that's give a round of applause to Marshall. Thank you very much for Marshall. Feel free to come and ask and yeah, Yeah, rich, way have all Been here tonight.