TechCrunch 40, Day 1
I’m hanging out at the event of the day, the TC40, where Jason, Michael, and Heather have done an amazing job of getting sizzling hot startups to fight for a seat at the podium. Well, that’s the story, anyway. As if this event needs blog coverage, lol!
Update: On the off-chance you’re wondering, I never did get around to posting on the Day 2 companies, for various reasons. But seriously, the TC40 got an amazing amount of blogosphere attention, so it’s not as if more comments are needed.
As I’m watching the presentations and checking out the demos, I’ll try and rate them. My rating is from a perspective of “interesting”, in the sense of showing something that I think is really compelling – as opposed to something that’s an easy tech flip.
(Brief note from the demo pit: companies to watch from today’s set include BigSwerve, Stixy, and Tangler.)
Final version; I’m going home to bed. It’s been a fun day.
Session 1: Search and Discovery
There’s not much room for a new search company in the world, despite what the enthusiasts say. But conversely, it is a good space for tech flips, since there are buyers with deep pockets looking to continuously improve the quality of the search experience.
“Powerset is a search engine that focuses on natural language processing. The technology is based on breakthrough technologies licensed from PARC and developed internally, natural language processing allows users to express their intent in plain English. Using technology that understands of the structure and nuance of ordinary language, Powerset’s index is built by carefully reading and encoding the meaning of each sentence on the Web. By matching the meaning of a user’s query to the meaning in our index, Powerset is aiming to change the way we search online.”
The problem with NLS is that it looks extremely well in “toy benchmarks” – small examples. But scaling it to the web is a really tricky thing. There is indeed a lot of interesting technology in NLP (natural language processing), but historically it’s been very context-sensitive and fairly hit and miss. They have introduced some crowdsourcing techniques to scale better, but I’m immensely skeptical. Sure, any established search company might buy their tech to flesh out some of their current strategies, but NLS won’t change the search landscape anytime soon.
“Cognitive Code makes tools and systems for enabling practical conversational artificial intelligence applications, so users can interact better with their computers and mobile devices. Their SILVIA (Symbolically Isolated, Linguistically Variable, Intelligence Algorithms) platform is a complete system for the development and deployment of intelligent applications to almost any platform, with a technological core that allows humans to interact with computers in completely natural and intuitive ways. The platform helps derive context and meaning from user inputs, via speech, text, or other methods, so you can communicate with the platform as if it were another person.”
“Artificial intelligence for everyday problems.” Their main product is the SILVIA platform. Again, this is a technology that has worked well for a long time in demo mode. (This is the kind of technology that makes you insane when some company buys it for their call center.) Their initial monetization is focused on toys.
Similar to NLS, it’s yet another application of high-end software techniques; which is nice, and may be a useful technology/talent asset for existing net players, but not compelling per se.
When challenged about this, the founders had two comments: first, computing costs have fallen to interesting level, and secondly, with search, there’s a monetization option. To which I say, yes, true, and Google already knows this and is busy putting a lot of AI into their search engine – as are other search providers.
“CastTV is trying to build one of the web’s best video search engines. CastTV lets users find all their favorite online videos, from TV shows to movies to the latest celebrity, sports, news, and viral Internet videos. The company’s proprietary technology addresses two main video search challenges: finding and cataloging videos from the web and delivering relevant video results to users.”
Maybe it’s just me, maybe I’m too old, but I don’t quite get this. The problem with searching video is sorting out what’s good. YouTube is fun for a while, but mostly it becomes a distribution platform for videos (like key scenes from network news) that are referenced from other sites. So YouTube’s real strength today is branding.
That said, CastTV showed some pretty cool search technology. It’s pretty clear that the current search engines are behind the curve on best practice. Like VentureBeat notes, companies like CastTV are good tech purchases.
“FAROO is a peer-to-peer web search engine that has no centralized index and crawler. Each web page visited by users is automatically included into the distributed index. Ranking of search results is based on a distributed usage statistics of the web pages visited by FAROO users, which leads to a more democratic, user centric ranking. FAROO also shares advertising revenues up to fifty percent with its users. The search engine uses privacy-protected behavioral targeting to increase conversion rates.”
Essentially, Peer-to-peer search engine. Without using a traditional crawler, Faroo’s idea is to index only pages visited by one of the P2P users. Instead of expensive, centralized resources, Faroo will scale with users, and also leverage user behavior to gather information on usefulness etc.
This was clearly the most compelling company in the search category. There are some alluring aspects to P2P search. But there are a number of issues as well. Firstly, it presents a chicken-and-egg problem: what level of usage
“Viewdle is a media platform for indexing, searching and monetizing video. The technology they are developing will let video content owners extract metadata from news, shows, movies, and Internet video. Content owners can leverage both their new and old content with Viewdle’s internal and white-label search and indexing capabilities to maximize relevance, usage, audience and monetization. This is much more effective than the old method of text-based metadata indexing. Viewdle has a killer feature in its facial-recognition technology. It is able to index video frame-by-frame and create a “real-time index of true on-screen appearances with unrivaled accuracy and relevance.” They plan on building one of the largest databases of people-in-video references. Reuters is currently testing out Viewdle’s technology with their videos news inventory by letting people search their catalogue for specific people.”
Ok, so, NSA will love this, but I just find this scary.
And, yet another Britney Spears example. It seems that 90% of searches on the planet are either Britney Spears or Paris Hilton.
Viewdle is a real good reason for not sharing your personal videos on any indexable location.
Session 2: Mobile and Communications
Anything that relies in any way on collaboration with mobile operators is a challenge; the only player that stood out for me in this category was Ceedo.
“Cubic Telecom is a mobile service provider looking to become a fully global mobile startup by blanketing the globe with their signal. The company aims to drastically reduce international mobile roaming and call charges, which will lead to cheaper prices for overseas calls. They will accomplish this goal by utilizing existing phone network resources without the need of downloads or new phone configurations. Founder Pat Phelan a well known communications blogger “wants a world in which anyone can pick up their mobile phone wherever they are and call anyone in any country for as long as they like without worrying about the price.”
The “pre-game chatter” had this company pegged as a favorite. 500 million international travelers spend $20B+ on roaming charges when traveling. Check out Max Roam.
They’re dubbing themselves the Ryan Air of international cell phone. The notion is to sort out various issues with allowing a single SIM card to programmed for local phones in up to 50 countries (e.g. you will have up to 50 phone numbers). They’ve fleshed out the solution to include mobile phone unlocking services etc.
But the challenge is, how do you build a sustainable business model? If anything, this is yet another reason to short cell phone operators, as if we needed more. If we spend $20B+ on roaming charges, and they take the example of a $1400 bill from one trip to France; well, what if I want a 99% savings to use their service? So that’s a 200M market. And I have alternatives – callback services, picking up a rental phone, having a second (or third) account for my most-travelled-to market, etc. When I last travelled back to Sweden, I picked up a cash GSM at the kiosk at the airport, and in 10 minutes had a new number, I called my US phone number and changed the voice message to give my temporary traveling number. Done.
“Yap provides voice-to-text translation services for mobile phones. Users can say anything they like and Yap will send a text copy to anyone of their contacts. The service is completely automated so you won’t have intermediary Yap employees listening to your messages, typing them and then sending them out. They also have a text messaging application call Yap9 that allows you to keep in touch with friends, family, and co-workers. Users can also use the application to instantly query mobile web services just by talking. They can search Google, Wikipedia, Yahoo, and YouTube, or interact with Facebook without using their phones’ miniature keyboards.”
These guys garnered sympathies when they had a serious A/V issues. Jason loved these guys and jumped to their defense – “when you see it, it’s beyond cool”.
Yap tries to create a full voice-governed experience of mobile. It looks very good, and again, seems eminently acquirable as a tech play.
This company seems to be generating a lot of interest.
“Trutap is a mobile social-networking application that enables users to stay connected with offline and online friends regardless of where they are in the world. The application also claims to work with all social networks, IM clients, network carriers and mobile phone devices. Trutap is accessible through a downloadable mobile application and their web browser experience synchronizes all Trutap messages, conversations and contacts. The application lets users communicate with individuals or groups via text, picture and instant messaging. The service also enables users to link to their social networks, upload pictures and update their blogs.”
Oops, I was busy responding to comments on my posting yesterday about Jane Fonda and nuclear panel, so I missed the demo. But they beatboxed so they must be cool!
“Ceedo Technologies is a virtualization software company headquartered in Israel. Its patent pending approach to virtualizing the Windows’ desktop environment enables users to carry their PC-based work environment on portable devices such as USB flash drives, pocket hard drives, network drives and even mobile phones. Ceedo works well with the mobile device market. This is because it does not virtualize operating systems, which lets it load and operate more quickly while taking less drive space. Ceedo Mobile technology lets users connect their favorite mobile device to a PC without requiring installation or configuration.”
Virtualization for smart mobile devices. Launches on a PC from an application stored on the phone, and can then launch applications from that mobile phone and run them on the PC. The demo included editing Picasa pictures, and posting them to the blog. It blends understanding of what phone you have with the PC; thus if you buy music or video, it’s selected such that it’s viewable on the phone.
It’s a very clever packaging of virtualization technology, and there’s an obvious buyer with strong share price out there, so these guys should do well.
But it doesn’t work with Macs (sigh) …
“Loudtalks is a free downloadable push-to-talk messaging application. It supports real time private or group voice communication in the walkie talkie style. Loudtalks is based on the peer-to-peer architecture and capable to work behind most firewalls and NATs. The software is lightweight (installer is less than 1 Mb) and unobtrusive. Normally it runs in the background and can be activated with hotkey without switching the focus or popping up the application window. The advantages of Loudtalks over existing messaging systems include speed, asynchronous style and voice messaging.”
“Push to talk for the internet.”
Not much different from gaming audio platforms like Ventrilo. Except those work quite well, whereas Loudtalks also had serious problems demonstrating.
Session 3: Community and Collaboration
Storyblender (Enfra Networks)
“Story Blender is an online collaborative video production platform where people can work together to “blend” their media for rich, interactive storytelling. Story Blender CEO Hyoung Yong Joon is the original founder of Cyworld, which is South Korea’s first and most popular social networking community. StoryBlend’s online editing tool lets users create videos by “blending” images, sound, text, and video clips. When users have created new video blends they can then share it with their friends and the StoryBlend community.”
They draw an analogy with mediawiki. The CEO is presented as the “original founder” of CyWorld – I thought Cyworld was a big corporate project (SK Group)? Anyway, I don’t quite get online collaborative video editing. And when pressed on the hot issue of copyright, there was a confused answer.
“TripIt is an intelligent travel organizer that helps do-it-yourself travelers manage their travel plans so that their trips go more smoothly. Travelers simply forward their purchase confirmation emails to TripIt and TripIt automatically creates master itineraries with travel plans and other critical information like weather, maps and driving directions, and destination information. TripIt makes it easy for travelers to print or access their trip plans from anywhere including online, in print and on their web-enabled mobile devices. They can also share itineraries and travel calendars and collaborate on planning trips with friends in their TripIt network.”
“We’re not in the travel business, we’re in the information management business.”
Tripit has a brilliant and simple idea: regardless of what travel-related web sites you use, whenever you book something online (hotel, plane, limo) you simply forward the details to a magic email address (firstname.lastname@example.org) and it will intelligently parse the information, complement it with interesting travel-related information, and provide an environment for keeping sharing, printing, etc.
This is an excellent application of NLP/AI/semantic web etc. It’s also a real problem area. And travel is an excellent monetizing target.
This company is the one to beat so far.
“Flock is a social web browser. When using Flock, people can easily discover, access, create and share videos, photos, blogs, feeds and comments across social communities, media providers, and popular websites. Flock is also a powerful platform for distribution and marketing partners who receive a custom browser optimized for their business to drive engagement and incremental revenue from their communities both on and beyond their site. To date, Flock has shipped editions of its browser for Photobucket and Piczo.”
Based on the Mozilla web browser, Flock strives to make Mozilla into an excellent social networking environment. They today announced version 1.0, so they’re going one-dot-oh.
Flock looks like it’s getting really interesting. I’m not sure how it’s to be monetized, or how to hold up against open source alternatives being rolled into Mozilla. But it’s worth watching.
“MusicShake is a Korea-based online music creation service developed and distributed by SilentMusicBand Corp that provides music composing solutions aimed at the general public without previous musical knowledge or expertise. The service lets users create personalized, professional quality music using various tools and pattern-combination methods. They hope to meet the growing demand for customized ringtones and personalized music that mobile phone users and Internet users have for adding personalized creativity to their personal blogs, websites and social network pages.”
This is an application, basically aiming to topple software like Garageband in making it simpler to put together simple music. Not sure what the business model is, even if they have a slide saying “We have lots of ‘REAL’ business models”. It looks like cool software, but packaged software is packaged software.
“8020 Publishing is a media company that allows online communities to take an active role in creating 8020 Publishing’s published print magazines. They currently have two magazines JPG and the yet-to-launch Everywhere. They let their online users handle the heavy role of creating the content. They also let them critique and vote on the print content. However, 8020 Publishing still fills normal publishing roles like choosing themes, putting the magazines together and providing the final vote on all published content. Their editors act more like curators letting their contributors the online communities provide the raw content. Taking this approach has provided them with a built-in magazine subscription audience not to mention loyal online communities. It has even led to them receiving two industry awards in 2007, Circulation Innovation of the Year and Best Use of Viral Marketing for Audience Development.”
One of Jason’s favorites. “Community created magazines” is very, very interesting. They are seeking to combine the benefits of web collaboration with the magazine structure. They launched JPG a year ago that went well; today they are announced Everywhere Magazine, which will launch in a few weeks.
There’s is a middle ground somewhere between blogs and printed newspapers. 8020 seems to be exploring it in a very focused manner.
Session 4: Crowd Sourcing
“Cake Financial is a social investment service that lets people safely and securely track all their investment portfolios in one place. The service allows individual investors to track and analyze their historical performance up to ten years. Users can also view the real-time portfolios and performances of their friends, family and top investors all without disclosing net worths, shares owned, portfolio sizes, etc.”
“You can’t game the system.” Hm. Ok, I admit, I don’t get it. I don’t see friends and family sharing this sort of information to this extent. And if it’s strangers, it’s back to the monkeys-and-WSJ: if you have a few hundred throwing darts, some of them will do a good job of picking stocks. And what is your interest as an individual contributor. The panel gave Cake a hard time.
“Docstoc is an online community and professional network geared around user generated, professional documents. The site integrates the most useful features of private online file storage with the ability to store, categorize, and share content found anywhere on the web. Users can search for documents by categories or by keywords and then preview the documents online and download or store any content for free. They can filter search results by views, downloads, ratings and comments.”
Nice thought, but where do the documents magically come from? Tags on Google search help me find various documents on the Web, and has for years. So is this just a semantic filter? I think it’s a great concept, but getting the squirrel wheel spinning can be tricky; they might want to focus on one topic (vertical) to start.
“Teach The People is a social network built around online education. The site lets anyone with specific subject knowledge or a useful skill set share with the Teach The People communities. Users get to create individual profiles and contribute content to topics such as computer programming, math or “Bob Marley’s Influence on R&B Music”. The site encourages quality content by letting users become community creators and by giving users points for rating, referring friends and answering questions. Community creators help create content and run day-to-day community operations. They can charge other users fees for monthly community access, content views or content downloads. They can also share in site advertising revenues.”
These sorts of “fat verticals” I think are the model for finding good revenue sources.
“CrowdSpirit is a crowdsourcing community built around designing electronic products and staying involved throughout their product life cycle. Users submit ideas for innovative electronic products that the community fine tunes and votes on. The best ideas and their product specifications rise to the top where investors provide financing and development partners make prototypes. Once products have been made they are tested by the community and recommended to retailers. Users involved with product creation can earn a share of the product revenue. Typical products will include MP4 players, DVD players, computer peripherals, headphones, etc.”
There’s a fine line between effective crowdsourcing and design-by-committee. Is this project management software? I don’t see it.
Designing a product and bringing it to market is extremely difficult to do well. The examples cited are extremely competitive product segments.
“The world’s first personal manufacturing platform, Ponoko is the online space for a community of creators and consumers to use a distributed network of digital manufacturing hardware to co-create, make and trade individualized product ideas, on demand.”
Very cool, very clever, but I don’t see it becoming big.
The panel ended with Yossi Vardi quoting from Theodore Roosevelt’s speech from 1910:
It is not the critic who counts; not the man who points out how the strong man stumbles, or where the doer of deeds could have done them better. The credit belongs to the man who is actually in the arena, whose face is marred by dust and sweat and blood; who strives valiantly; who errs, who comes short again and again, because there is no effort without error and shortcoming; but who does actually strive to do the deeds; who knows great enthusiasms, the great devotions; who spends himself in a worthy cause; who at the best knows in the end the triumph of high achievement, and who at the worst, if he fails, at least fails while daring greatly, so that his place shall never be with those cold and timid souls who neither know victory nor defeat.
An appropriate wrap to the presentations! Good luck to each and every one.
- Posted in: TechCrunch20 ♦ TechCrunch40
Thanks for the profile. What makes SILVIA unique amonst AI is that she is context sensitive as compared to the current versions of AI. She is actually quite different than anything else out there and the people who understand the difficulties in getting an AI to respond (see Ray Kurzweil’s Ramona for instance) are all blown away. There are teams of AI research in every single major institution (MIT, Princeton, etc.) in the US who have been trying to solve this for years. We’ve never seen anyone else have a conversation with a computer, unless you speak in specific command phrases. Can you have a conversation with Google? I think not.
Peter, you are a very good innovation journalist!!
[PSM] Why thank you! 🙂