Transcript of Moxie Podcast 4 (on crowd-sourcing)

Courtesy of the kind folks at Transcribe Me, whose Alexei Dunayev features, here is a transcript of Moxie Podcast Number 4.

You can, as  always, listen or subscribe to the series via iTunes or follow along with the audio via Mixcloud.

Glenn 00:00 Hi there, my name is Glenn Williams, and welcome to the Moxie Podcast, episode number four. I am with you here in London, and very shortly I will be bringing my guest on New Zealand now if you are coming along to the Moxie  podcast for the very first time; you might be wondering what the heck this is. Especially if you are outside of Auckland you would not have had a chance to go along to the Moxie Session, which happens about once a month in Auckland, and basically it is a bunch of like minded or may be even disagreeing people coming together to discuss a different topic each month, but essentially talking about retooling our existing roles and attitudes for the internet and making connections that are usually missed or ignored between different economic areas. But essentially it is the internet that is bringing us all together, the internet economy. Now the topic of discussion at the last Moxie session was; is Crowd Sourcing, one of the next big things for business and government, when it makes sense, when it does not and who is actually doing interesting things with sourcing?  And with you can actually rely on the crowd for the heart of your business. That will bring on the three speakers for this topic. The first being Lillian Grace, now Lillian is the founder and chief of Wiki New Zealand, which is a collaborate website making data about New Zealand accessible for everyone. She will so consult with Stakeholder Strategies. A firm that works with public and private organizations and NGO’s on commercial, economic and social-environmental issues. She is also previously been at the think tank New Zealand Institute Lillian welcome to the Moxie podcast.
Lillian 01:42 Thank you Glenn.
Glenn 01:43 Doctor John Small is the second speaker and he is an economist working on firm and industry issues. He was an academic economist for eleven years and has been consulting to industry and government clients since 1991. He works through COVEC and Open Source firm he founded in 2001 and took about things he co-founded or founded, he has three failed Internet start-ups to his name. In 1994, 2001, 2007 is one thing we have learnt from previous Moxie sessions it is important to fail. Dr. Johns more welcome to this Moxie podcast.
John 02:22 Good morning Glenn and thanks for coming.
Glenn 02:25 And Alexei Dunayev is our third and final guest, he is an early startup technology CEO and serial entrepreneur, he has a reputation for excellence and Innovation creating winning businesses through a market niche, passionate for online education and e-mentoring. Here is the CEO and Co-Founder of TranscribeMe, let us say, welcome you to the show.
Alexei 02:49 Good morning Glenn, thanks for having me.
Glenn 02:51 And pretty much, I can sort of already tell that you guys are very interested in this topic of crowd sourcing and through your business and daily lives deal with crowd sourcing in one way or another, but Lillian first I will get you to talk through briefly your points at the last Moxie session. Because you are talking about the unexpected benefits of crowd sourcing, almost the side effects I suppose.
Lillian 03:15 Well I think that Crowd Sourcing that really valuable approach that we are really at the beginning of understanding, but I do not think that should be  at the heart of organizations and businesses and that is kind of what I talked about at the Moxie session.  So with crowds I think there are lots of different benefits we can get from being able to mine collectible intelligence from groups and use wider pools of skills and expertise to contribute to our efforts, and I think there is a huge potential for using it kind of internally in large businesses and things. In terms of being able to break down normal power structures by using crowd sourcing tools, but, despite all the benefits, which I can talk to more in detail. The reason I do not think we should rely on this for the heart of most of our businesses and organizations. Not for all of them. Is that although people love to contribute, their focus is generally elsewhere. The more engagement you ask for, the less contributors you get and I think the most important factor in a successful venture is to have a really strong vision and to have somebody responsible for executing on it. But crowds do not feel responsible and we know this from you know science studies that have been done showing people in the waiting room with smoke out the door and no one reacts because they assume someone else to be responsible. So I think we need to engage crowds in ways that outcomes are not reliant on them and that we still need to retain like the core team and the core focus for most of the businesses and organizations with the exception of those companies said that exist because of crowd sourcing benefits.
Glenn 04:52 Alexei your business, TranscribeMe that is a service to provide fast accurate and cheap audio to text transcription. That crowd sourcing is very much the heart of that business, can you give us a summary of your talking points from the Moxie Session from your point of view? …
Alexei 05:13 Yeah, thanks Glenn, so as you said TranscribeMe provides a very accurate, very fast voice to text recognition service and we do that through a hybrid model that combines speech recognition algorithms with real human crowd sourced transcribers. And the reason that we engage the crowd in this venture is simply because there was no other way to achieve our mission and our goal of accurately and quickly converting voice to text. If you look at the state of computer algorithms alone, they are very fast and right now they are very cheap, but they are no where near 100% accuracy that is required by a very large subset of people in our customer base. So we need to supplement our computers with people and if we use individuals working in an office under fluorescent lights, there is only so much you can do, there is only so much capacity that you can create. However, if we have a new model that allows people to work at their convenience and allows anybody in the world to monetize the smallest possible amount of their downtime while contributing to productive work. I think we have got an ability to really create a service that is scalable as well as accurate and fast. So to give you an example, our crowd consist of people that work whenever they are free so we could have transcribers that are sitting on the couch, they are watching TV, and during the ad break they may go and do a few minutes of work on the TranscribeMe portal and then they go back to what they were doing before. This is really at the heart of the TranscribeMe model. We are able to bring people into our crowd by giving them an opportunity to monetize their unused downtime. And so that also provides us, not only with capacity, but provides us with level of accuracy that cannot be match by computer alone, and one of other things we do a little bit different to what Lillian has described is that our crowd is very much working on an individual basis.  In the sense that nobody in the crowd has a clear understanding of everything that is going on in the system as opposed to a common website that people contribute to so this is actually quite deliberate because we want to maintain confidentiality in our service. And by giving our transcribers only very small, 10 seconds, 30 seconds pieces of audio, we do not give them the entire context of the recording and they are almost deliberately kept in the dark about the reminder of the work. To maintain confidentiality, so that is a little bit about the crowd and what we do with it.
Glenn 08:00 That is quite a specific example of Crowdsourcing. John, at the Moxie Session, you had a broader look at the variety of things that Crowd Sourcing could be useful for. Could you give us a rundown on that?
John 08:16 Yeah sure, I started to think about, this is a, in terms of the range of different activities that could fit into a business or an organization. I think they are some strong parallels between not all but many of the crowd sourcing models and what we know is platform businesses. And the platform business is one that there are an agents that mutual benefits it. You think about for example a TradeMe is a platform business that brings buyers and sellers together. And any kind a trading platform or connectivity type of platform even a payment system or any of those sorts of things, these are all platform businesses. They have some specific economic characteristics and one for example is that they tend to be quite hard to get started because you need to have buyers if you want to attract sellers and vice versa. So it is quite hard to get cracking, but once they do get cracking, they are subject to these quite strong network effects that that make them really, quite powerful and stable operations. Now, I am not saying that all crowd sourcing applications fit into this model; a number of them. The second thing I went through was, a bit of a breakdown of the kinds of things that can be crowd sourced and are not crowd sourced. I was drawing on some early work, the plan is this sort of thinking, and classifying this into four different things that might be sourced. One would be wisdom or skill, and you could think about production markets as is doing that. Certainly, Alexei’s business does this and there is a number of others out there like Amazon’s Mechanical Turk. So that is one thing that could be obtained from a crowd, the second one would be content, things like TripAdvisor and Wikipedia, which is all user-generated content and we know the value of that and many of us read that information and fewer of us contribute to it. The thing is voting, and that could be as simple as polling, regular polling businesses that are essentially Crowd Sourcing opinions but there are other interesting applications as well.  And then the fourth thing was funding, Crowd Source through things like Kick Starter and there is a journalism one called Spartax. So those are the things that you might get from a Crowd and you might plug those into a standard kind of business. I think we all, and just a thought a little bit about the sorts of challenges that, what you would face in a Crowd Sourcing, and really it seemed like there were two big challenges. One is what else, being very careful and very clear about exactly what it is you want in the Crowd to supply. So, that is a design issue it comes from different ways and different businesses. Say for example polling [inaudible] they are concerned about keeping buyers out of this and polling and other questions. For things like fundraising you have an issue about specifiying the contract, exactly what are you are getting into when you put money into one of these start ups. And then there are other things you know like reputation systems and we have people who give feedback on the interactions they had with others. And in most systems it is quite important to try and foster honesty. So that is one sort of challenges just designing the task quite carefully. The second set, and Lillian certainly touched on this, and it is key to Alexei’s business as well is how do you motivate the Crowds? Why are people helping you? Why does it matter to them? Is there some sort of common purpose they are getting out of it? Or, perhaps there is some financial incentive that needs to be provided?
Glenn 12:35 That is a good point to pick up on as well sort of Crowd Sourcing I suppose can lend itself to self-selection, or to be populist in some ways, is there a danger – I will throw this out to the floor, if anyone who wants to grab a hold of this – is there a danger that whatever your project is, whether it be creative or governmental or whatever, is there a danger that you could be lead by the Crowd rather than being innovative with an out of the box idea? Is that a danger at all?
Lillian 13:11 I think it depends on how you structure your organization because back to my earlier point, to have someone with a strong vision and someone that is responsible for it. If they are really strong and– like I described it, when you try and engage the crowd I think you should try to engage them from the moving train. By that I mean do not stand on the station calling to ask the crowd to help you try to get on the train or what direction to take. But actually know where you are going, be on the train, be moving and then tell people where you are going. Articulate it in a way that they feel like you are inviting them on board to say we are going here and we would love your help. We would love to have you on board and I think it is definitely a danger of that happening, depending on how an organization is run. But if the focus is kept and if you are engaging people into that you can definitely structure I think in ways that and they get crowd benefits to help you and maybe tweak and change tact, as you get good ideas from the crowd, but not necessarily just changing what you are actually doing.
Alexei 14:17 I can probably add a little bit to this. I think that there is a lot of value to be gained from engaging the crowd beyond the basic tasks that they have signed up for, and in TranscribeMe we have a social network that we developed on the back of Yammer specifically for our crowd workers. Where they are not only able to talk among each other and share ideas of how to be more efficient and how to do more with the resources they have, but they are also able to give us suggestions and give us insights into ways we can improve the systems and improve technology and that really does go, I think, above and beyond their call of duty, so to say. So there is certainly an opportunity, in my view, for listening to the crowd beyond what they are tasked with. However, there is the common saying that your customers can tell you what you did wrong, but they can not tell you how to fix it and that certainly applies in product development. I think with the crowd it’s sort of the reverse of that coin, the crowd can tell you how to fix something because they do have that day to day experience of engaging with your systems, with your processes, with your content. So both listening to the customers and listening to the Crowd. I think it is equally important and fulfills different roles in terms of gathering knowledge inside the organization.
John 15:48 You know what, I agree with that, but I just, it would also come back to Glenn’s point which is a fear that the Crowd can lead you astray. Now you think you could the crowd could get off in an predictable direction and turned into a mob. You know we talked I think briefly last week about the–
Lillian 16:13 Yes.
John 16:14 Mob sourcing. There are certain examples of, even sort of malicious reorienting of crowds, I think the finest one with Justin Beiber was sent to North Korea, but his fans, his alleged fans, that were not really his fans [chuckles].
John 16:32 Yeah.
Alexei 16:32 So yeah.
Glenn 16:33 Well, most criminals are probably still being hanged I guess if law was Crowd Sourced, right?
John 16:40 Exactly, you have a point.
Glenn 16:43 Someone just brought up the issue of perhaps internal Crowd Sourcing within  organizations and internal communication. I know this was discussed at the Moxie session moving away from a hierarchical expert models of organization. Is that really feasible? Can you crowd source within an organization to take away some of the tough decisions from managers?
Lillian 17:09 I do not think you can take away tough decisions necessarily. I think that you can get huge benefits from really engaging people throughout an organization, getting good ideas from people that are kind of at the face of customers and dealing with issues on the ground. But I personally do not think that you could use the Crowd or the internal– the whole organization to make some hard decisions. I think there are some decisions that nobody would choose to make individually but when your responsible for the overall  outcome you would make differently
John 17:50 I agree with that crowds are great for making easy decisions like what should we name this product and things like that but, should we close this plant and move it somewhere else, that is a tough one.
Lillian 18:03 Yeah.
Glenn 18:04 I guess everyone needs to have all the information in hand. Is this one of those we overcome with Crowd Sourcing in general? Is being able to make complex information available so that people can make a decision?
Alexei 18:20 That is an interesting question, Glenn. I think that I am probably of the same viewpoint as we were just been talking about that it is hard to make a decision without having that information. But I am also advocating that decisions like that as should not be made by a group or by a committee, there are definitely times in a company’s life cycle and in any decision making life cycle when there needs to be one person who is responsible and in charge of making a decision. So, sort of management by committee has never been a personal favorite of mine and I think that extends to management, decision making as a crowd also. Ultimately, I think when it comes downs to is that it is very good to draw on the knowledge and skills of the crowd to develop options and so in that divergent phase of decision making when you really try to maximize the different options that could be achieved. Then as this decision making starts to converge towards a single decision it is really the case of having somebody who is responsible, has got all of the information, and is able to remove the options that are not feasible and to narrow down on the one that is. So I think it is really a different phases of decision making that the crowd plays a role in.
Glenn 19:47 In the discussion at the Moxie Session, did any of you come up with your dream scenarios with Crowd Sourcing, should or would or could take place and if not you are able to come up with any now, we would really like to see Crowd Sourcing implemented?
John 20:04 I have a couple of comments on this at the Moxie one was that, and this is coming at from a societal angle, one was the observation that many of our problems– many of our sort of best ways of solving problems start by asking the right questions. I suggested that it might fun to try, and Crowd Source useful questions that would be would be relevant to the way our society’s organized. We did also, actually, touch briefly on sort of the issues of democracy and how you might make public decisions?  There was some discussion about that and I think it is relevant to the one ones we were just talking about regarding the extend of public decision making can be devoted to the Crowd. For example in theory you could force so you can have Crowd Sourced laws, Crowd Sourced justice and all that sort of stuff. That gets to the point of saying, how confident are we that, that would result in good outcomes and it builds some tension in the group between people who thought well I could get there and people who thought no you need to have actually a strong decision maker somewhere in there who is responsible for filtering all this stuff and making the right decision.
Glenn 21:39 Anyone else there with any thoughts on where they would love to see Crowd Sourcing, or where it is even just taking off right now, and it is in it’s infancy.
Alexei 21:50 Yeah, sure, Glenn, and I will probably be a little bit more specific with the example this is not happening yet but if you look at professional services, if you look at lawyers, if you look at accountants if you look at lawyers, if you look at accountants, if you look at consultants. That is all work that is sort of provided by a very structured hierarchy and sort of a single person somewhere in it, does the job. But if you draw the parallel to transcription to translation, which are already crowd sourced, those fields are essentially entry level professional services such as transcribers, typically somebody who has got a 2 year college degree and specializes in particular type of transcription, say medical. Same with translators, and if you sort of extrapolate that line to other professional services, if you look at let’s say paralegal, if you look at the entry level accounting services. These are things that are really next in the pipeline for getting Crowd Sourced. To give a specific example, imagine if you are somebody who needs to have a 100 page legal document reviewed. Now you could have 1 person go through the hundred pages or you could parcel it out and have a hundred people each look at one page and then have somebody review their findings and you can imagine which process would yield a faster result, providing that all the accuracy checks are in place, of course. So, if you look at the professional services field, I think that as a ride from disruption with crowd sourcing in the future.
Glenn 23:25 Of course scientists have been implementing this for years have they not? With the whole peer review processes of scientific papers. So they have been on to this one for quite a long time.
Alexei 23:35 Crowd sourcing is certainly not new.
Lillian 23:37 Yes, it is slightly different, but I totally agree with the examples that both John and Alexei have given. But, just thinking about Wiki New Zealand and also going back to your point about with crowd sourcing, does that mean we need to educate people to be making good decisions and to not be using good input into what you are trying to crowd source for and if you think about New Zealand as a whole, and think that really we Crowd Source our overall performance in a way if you think about it like our overall economic performance, our overall societal performance, our overall environmental performance all come from – not entirely – but they come largely from the actions of individuals through out New Zealand. So with Wiki New Zealand the whole purpose is to try and educate and to lift the level of informedness throughout New Zealand so that people are making better decisions in a way trying to educate the Crowd that leads to New Zealand’s performance and whilst bearing in mind of course that is not the only indicator of that performance.
Glenn 24:41 It is a good point, when we talking about Wiki New Zealand, as well as part of the effort of Wiki New Zealand, also, to make sure that institutional knowledge – institutional being the Institution of New Zealand – is also not lost over time and we do not make the same mistakes over time because we have got all this knowledge to refer to from the crowd.
Lillian 25:04 Yes, all the content on Wiki New Zealand though is in data form like I kinda see– I think data is a really, really important part of making decisions, but I acknowledge, that it is only the first step in the pipeline that goes through knowledge and wisdom.
Glenn 25:22 Yeah, then the difference is data is now readily available as opposed to 10, 20 years ago, we have to go and dig through papers and screens of all computer systems and in far flung corners of New Zealand, right?
Lillian 25:33 Exactly and it is currently throughout the world. It seems that most data is presented in either databases or kind of embedded in spreadsheets and so to use it, which I would like you to be visually, to kind of understand it, you need to have time both time and skill to access it and right now when we see data in usable forms, it is generally because somebody is trying to convince us of something, which is a very legitimate use of data, of course. But in terms of understanding how we have performed over time, like New Zealand’s history and managing where we are now as it relates to vision where we might go. I think it is really important to understand and to see some of the data that people are getting fascinated by with Wiki New Zealand is unusual things like meat consumption per capita. For example, between 1970 and today chicken consumption has gone up six times and beef and lamb have both halved. It is just things are there that you don’t know our country because why would someone do that?  No one wants to convince you anything about it. So that does not exist. So, yeah definitely one of our purposes is to, is to make people aware of our history and where we have been and not to loose understanding of what they past has been like. And that what we have today is not normal, like it is normal for us today but it is not, we can not expect it to continue.
Glenn 27:02 It is also going to be interesting to see how data over the past few years has influenced the way some journalism has been done and how data is being used to the keep institutions or politicians or whoever– much more honest at the end of the day because it is now out there and it can be Crowd Sourced. A whole lot of people can pour over that same data and come up with the same conclusion.
Lillian 27:25 Yeah although my one concern with that is that I think when people have to operate out of fear. It is a horrible way to operate and it does not really elicit good outcomes. Often people wield data as a weapon and if someone gets something wrong they use it against them, and I hope to create or to help establish an environment where people can use data and understand that there are lots of ways of slicing and dicing data and you do make mistakes and that if we are all looking then we can point it out in a way that… Do you know what I mean, often when people come out and say one data point and then people look and everyone is having a big fight if you are wrong or this is right rather than everyone trying to create a shared understanding. Because, it is really black and white. Even though at the start of this we always look at unemployment. For example, using one or two different types of data points when there is many ways to slice and dice unemployment.
John 28:27 I agree with that Liliian, but do you not think there is something a little more fundamental in there rather than just what the data say. People do not want to accept information that conflict [inaudible] in a way it is inevitable, I fully signed up to what you are saying, it would be nice if we could have civilized conversations about what facts really are but I fear that actually the reality of society is that as the people have do very strong prior beliefs and they have a strong desire to validate those and that is what gets in the way of having that sort of effect on conversation.
Lillian 29:24 Yes, I agree and I do not at all this is going to be an easy journey but I think it is worthwhile trying and I think that a lot of what we all believe is based on what a few people who have had access to data and kind of their conclusions on what they share. Rather than necessarily what we all individually believe. We just hear two different points and pick one often. So, if we can all understand more and hopefully we can come up with more than two options or more than two arguments for every circumstance.
John 29:56 I agree, I agree.
Glenn 29:58 And put it all in a nice picture, so that the rest of us can understand as well. This podcast has been a nice journey and hopefully it has summarized what happened at the Moxie session last week. We have been with Lillian Grace. Lillian, where can we see you online and what your up to?
Lillian 30:20
Glenn 30:23 Also, you are on Twitter as well?
Lillian 30:26 Yes I am: gracefullillian.
Glenn 30:28 Excellent. Great. Dr. John Small, Dr. John small. What about you?
Alexei 30:32 Yes. I {inaudible]. I am refraining from twitter at this point.
Glenn 30:39 Really?
Lillian 30:40 [laughter]
Glenn 30:43 Let’s go, it is like the ultimate and social crowd sourcing social media crowd sourcing….
Alexei 30:49 Absolutely.
Lillian 30:50 Yeah.
Alexei 30:50 Absolutely is, absolutely.
Glenn 30:53 Alexei Dunayev, what about you?
John 30:54 Yes, you could read more about TranscribeMe at www. and you can find me on LinkedIn.
Glenn 31:03 On LinkedIn, fabulous. Well, thank you to you all for being part of this Moxie podcast episode number four and go along and check out their websites or indeed go along to the home of the Moxie podcast. It is the Thank you also to the convener of the Moxie sessions and the Moxie podcasts. He is the producer who puts it all together, make sure everyone turns up, I just need to put the microphone and push the code to the lining to the sliding glass is the name its put at all together. Thank you very much to Hayden. Once again Thank you to everyone for being part of this. Thank you for listening as well. Check it out on iTunes or up on Mixcloud. My name is Glenn Williams. Catch you next time.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s