Medical diagnosis company GeneSort, which specializes in personalized medicine in cancer and hereditary diseases, announced today (Tuesday) that it has been acquired by AID Partners for $23 million. The acquisition will provide the Israeli company with resources and additional support in order to expand into Southeast Asia and other global markets. The largest investor in the company is Moshe Hogeg’s Singulariteam investment fund, which invested $2.2 million in the company.
GeneSort specializes in genetic analysis of deep tissue and enables recognizing specific genetic mutations related to certain types of cancer. Oncologists can use the results of GeneSort’s diagnosis to customize treatments according to the specific genetic mutations of the patient.
GeneSort was founded with the aim of better understanding genetic DNA profiles of various patients and to help doctors select safe and efficient treatments according to the patient’s genetic profile. Some of you know that changes in the DNA sequence of certain genes can sometimes lead to changes in the normal functioning of the proteins generated by these sequences, which may eventually lead to cancer. The company uses sequencing technology that enables parallel sequencing of millions of DNA bases, and uses several tests (panels) that look for genomic changes in the patient’s tumor as well as changes of genetic background. The same panels contain specific detectors for genes that are clinically significant for treating and preventing the development of the cancer, and each panel tests the gene sequence and recognizes changes. Diagnosing each patient’s genetic profile, which is done via a biopsy and a saliva sample, allows the doctor to better plan the treatment of his/her patients.
Dr. Gil Pogozelich, Chairman and CEO of the company, says: “We are very excited by the acquisition and by the opportunities it creates for our company and for the technology we’ve developed. The acquisition provides us with opportunities to help more patients around the world. This is an important milestone for us, and also a validation of the strategy we are using. Our technology has the potential to help millions of people around the world, and we expect that the next efforts will create great value in the fight against cancer.”
Moshe Hogeg, Chairman of the Singulariteam investment fund which led the investment in GeneSort, spoke about the acquisition: “The technology GeneSort has created is different from those of other companies in the field. This technology not only has the potential to improve people’s lives, but also to save them. We are very proud that we supported GeneSort in the company’s early stages and we are looking forward to further development after the acquisition.” Hogeg also said, “I am pleased that Chinese investments in Israeli companies are bearing fruit.”
Work centers like WeWork, Merkspace and others have become very popular among entrepreneurs, start-ups, and freelancers from a variety of fields who work without a permanent office. Now, a new Israeli venture is trying to present something a bit different through a conference room complex, which operates according to the model of renting rooms at an hourly rate. The project, “Meet in Place,” wants to save you the hassle of reserving your office meeting room or the uncomfortable feeling of sitting in a coffee shop for 3 hours. The concept also offers the possibility of conducting meetings and conferences in the heart of the city.
A very clear purpose
“Meet in Place” offers a similar but different concept to the work areas we know, being a designated place for meetings and conferences in the heart of Tel Aviv – whether that be a professional interview, start-up work meetings or a presentation to be given over to 30 people. The complex includes 20 meeting rooms of different sizes and styles (including a classic meeting room, large meeting room, a Salon and a Grand Salon), all with WiFi connection.
A nice feature ‘Meet in Place’ offers is the ability to see the availability of the rooms at any given time (even a few months ahead, similar to hotel booking sites,), reserve them in advance and secure your reservation without even having to raise the phone. The whole process is done online through their website or application. Another perk is that meetings in an office space are considered tax-deductible, unlike a meeting that takes place in a café.
In terms of rates, the Classic Conference Room cost about 98 NIS per hour for up to 3 participants – each additional participant will cost an additional 8 NIS. The room can accommodate up to 10 participants. In the evening hours (starting at 19:00) the rate is lower and stands at 65 NIS per hour. The Grand Salon for up to 10 people costs 144 NIS for 3 people (each additional person will cost an additional 8 NIS) and the Large Conference Room that can accommodate up to 30 people costs 196 NIS for 3 people, and an additional 8 NIS per extra participant.
In addition, the rate of the Classic Conference Room includes espresso and soda water. Those of you who are not satisfied with espresso can order the Classic room, which also includes a cappuccino at NIS 108 per hour. The room with the highest upgrade, which includes cookies alongside the coffee costs about NIS 118 per hour. You can go through the rooms and see their exact specifications from the size of the TV screen to how many people it can accommodate. By the way, we checked and there is a ‘club membership’ that can be purchased, allowing users to buy and bank hours from 25, 50 or 100 hours at a time for a discounted rate of up to 20%.
On the surface, it seems to us that this is an interesting concept. For a reasonable amount, it is possible to have a quiet conversation over a cup of coffee with other people right in the center of Tel Aviv (the compound is located at 8 Rothschild Street). The brains behind the project are entrepreneurs Mickey Dorsman, who is also a partner at WeWork and Yaron Kopel, a former senior executive at SodaStream (so now you understand why there is a soda machine in each room).
Many people have never heard the name Scott Cook and would not recognize him if you saw him on the street. But you have certainly heard of the other, more recognizable figures of his generation, such as Bill Gates and Steve Jobs. Cook, who visited Israel last week, is regarded as one of the most distinguished and groundbreaking entrepreneurs in Silicon Valley and his impressive resume doesn’t fall short of those names mentioned above.
Intuit, which Cook founded in the early 1980s, may not have scaled to the heights of Microsoft or Apple. Thirty years later, however, it is considered one of the world’s leading and most innovative financial technology companies, with a Nasdaq market cap of nearly $32.5 billion.
In the early 1990s, after founding the company and leading it to sales in the hundreds of millions of dollars, Cook passed the reins of the company over to the next generation. He became a mentor and active investor in the Silicon Valley startup scene and was among the early investors in companies like Amazon, eBay, and Snapchat. He is also one of Alphabet (formerly Google) founder and CEO Larry Page’s close mentors.
Cook, who still plays an active role at Intuit (chairman of the executive committee), visited Israel this week and delivered an innovation workshop to Intuit development center employees in Israel. Intuit’s Israeli development center was founded three years ago as part of their acquisition of Check, and later Porticor. It is slated to grow substantially in the next two years to become one of the company’s strategic development centers, together with their U.S. and India offices.
All beginnings are difficult
Cook came up with the idea behind Intuit in 1983 because of his wife. She had complained to him about something fairly ordinary in the United States: bothersome home bookkeeping procedures. Although most people in Israel do not bother doing much beyond periodic checks of their bank account, in the U.S., managing home spending and income is taken for granted: Every resident and family is obligated to file a personal spending and income statement to the Internal Revenue Service at the end of each year. Depending on their circumstances, they have to pay additional taxes or get a tax refund.
Cook thought that his wife was probably not the only one regarding this as a burden. He started planning a product for her that would use a computer to make the paper-based bureaucratic process a little less frustrating.
Cook remembers, “I started designing the system by myself. Since the last time I did any programming was in school, however, I realized that I had to get help from someone who could do it a little better than I. I asked Tom Proulx, who studied with me at Stanford University, to join me, and we founded Intuit together.”
The road to success was not easy. The two founders were unable to raise money from investors for a long time. “It took us almost a year to get our first product to the market, and even then, I couldn’t afford to buy any luxury for at least three years. The first time I bought something was in 1987 – it was a disk player,” he says.
Quicken (the first bookkeeping program developed by Intuit) was one of the first software programs developed and sold to consumers. Marketing software was unknown at the time.
It is amazing to consider that Cook sold software to people who did not know what software was, and at a time when there was no money for marketing, nor any way to reach users through the Internet, which did not exist. When we ask him how he managed to do all this, he replies, “After we closed several distribution deals, we managed to reach several banks. They offered the product to their customers, and we got several dozen customers through them. These customers told their friends. By word of mouth, the number of users grew.”
The rest is history. After three difficult primary years, the company’s products now serve 42 million customers in North America, Europe, Australia, and Brazil.
“Behavior doesn’t lie”
What contributed to the success of the company’s product was listening to customers in order to understand their needs.
Long before Eric Ries and the lean startup movement, Cook and Proulx were among the first to use what we now know as “usability testing.” This means conducting tests in which potential users are allowed to use the product. In doing so, developers observe their user experience and discover things that may seem obvious to developers, but are not necessarily so clear to people who are going to use the product.
Cook says that, “After we developed the software, we showed it to friends and people not used to working on a computer or doing bookkeeping, and simply looked at what they were trying to do. We asked them to perform operations, and saw where they got into trouble. When there was something they couldn’t do, we went back to the code and redeveloped it until it was simple enough.”
What was the most surprising thing you found out from your users?
“For years, we did a user’s survey for various features in the system. At the end of every survey, there were a number of catalogue questions, such as gender, age, and place of residence. One of the things we asked in every questionnaire was, ‘Do you use the product at home or at work?’ For several years, we saw that almost 50% of our users answered that they use the product at work. We assumed, however, that since the product was designed for personal bookkeeping, they were probably bringing their accounts to work and using the software from work because there was a computer there, not at home. We simply ignored this question. Only after quite a few years did we catch on and decide to go into the question more deeply. We discovered that many users use Quicken for bookkeeping for their businesses. It was really odd for us, because the product was unsuitable for this, but many small business owners didn’t really understand how to cope with bookkeeping, and our product was simpler and more intuitive.”
Following this discovery, the company developed its second product, QuickBooks, which later became the most popular bookkeeping software in the U.S.
“This revelation was what led us to focus on user-based innovation. Instead of relying on speculation, we let the customers show us what they needed, what worked for them, and what didn’t work for them. We allowed the customers to try out the solutions we were developing out as early as possible, so that we could learn how to improve the product.”
There is no doubt that lean startup is an excellent way to work … for startups. How do you maintain this way of working in a company that has grown to thousands of employees?
“When we think about how to take our products to new countries, our engineers accompany businesses in those countries, and see how they really do things. We observe and learn from this situation, and the product reflects it. Instead of relying on theory and what we think we know about a specific country and its culture, they really get out there and research it.”
A successful venture for an investor
Cook was one of the early investors in Amazon, eBay, and Snapchat. When we asked him what he looks for when he invests in a startup, he sighed, and answered that investing in startups is very difficult.
“I wish I had a magic formula. Sometimes the idea is better than the entrepreneur, and sometimes the entrepreneur is better than the idea.”
With Snapchat, for example, Cook decided to invest after he saw a young and diligent student, even though he did not believe in the idea and thought it was terrible (“Who needs disappearing pictures?”). On the other hand, he invested in Facebook 13 years ago because he believed in the idea. “When I heard the story about an Internet site with pictures and notices about all the people in university, I thought to myself how glad I would have been had there been something like this when I studied at university. I heard that Mark moved to Palo Alto, and asked for a meeting with him. I liked the idea so much that I wanted to invest. I asked Mark to develop an internal social network like that for Intuit, so that everybody would know everybody else. Mark told me that they were a little busy right now.”
Do you think that there is something in common between Pierre (Omidyar, founder of eBay), Jeff (Bezos, founder of Amazon), Mark, and Evan? Is it something that you acquire and can learn, or is a person born with it?
“It’s a combination. Each of those people you mentioned is very talented. They can see things, opportunities, that ‘ordinary’ people don’t always see. In addition, these people don’t follow rules. They go on developing their idea, even if other people tell them that they’re crazy. But they also know how to listen to the surroundings.”
What is the most useful advice you ever gave someone?
“The advice I frequently give entrepreneurs at every stage is advice that I wish I had taken many years ago – to take a coach. If you’re a CEO or senior executive, one of the things you certainly won’t hear from your employees is the truth. The employees will tell you the good things that you’re doing, but not the bad things. They won’t take the chance of insulting you with the truth, so you definitely won’t hear about things you did badly. But you have to learn this. The coach can tell you what you did right and what you did wrong. I wish I had done that when I was a young CEO. When I did do it, I didn’t really get along with the first coach I worked with, so I decided not to take any other coaches. Some years later, I gave this advice to Sal Khan, the founder of Kahn Academy. He found a good coach, and that coach now works with me, too.”
What was the best business advice you ever got, and from whom?
“The most successful advice I got wasn’t necessarily given to me directly; I learned it by looking around and trying it out myself.” Cook says that at his company, they used to hold enrichment meetings for managers given mostly through presentations. In retrospect, however, it turned out that even a short time after that training, there were not many changes in behavior.
About 10 years ago, that same group of senior executives went for advanced training outside the office. It included a 45-minute practical exercise, and the rest of the workshop was lectures and presentations. A week later, the participants were asked what they had taken with them, and 70% of them referred to the 45 minutes of the exercise, meaning things that they did and practiced, not things that they saw. In other words, months of preparation for a presentation are not something that people remember; they remember what they actually did. “It shook up my world. I realized that I had wasted my time up until then,” he says. The conclusion is that people learn by doing, not by watching presentations. “That also connects with what I said about how decisions are made in a company: when there’s indecision about a certain question, the culture in the company is that it’s preferable to let the data speak: to do a trial and check the results, and make a decision on the basis of the results, not on the basis of a decision by some manager or other.”
The future of fintech
Intuit is almost 34 years old, a veteran company. The fintech industry is heating up, and we have seen quite a few revolutions in recent years: banks, insurance companies, everywhere people are talking about money. Now, people are talking about location, bitcoin, and electronic payments. How do you see us handling money 20 years from now?
“Today, various companies are interacting with each other frequently. The ‘real’ connection between them, however, still includes a paper invoice or mail and checks, for example. These transactions are still taking place like they did 20 and 50 years ago. All this will become automated, and will all meet somewhere in the middle with location-based technology and systems. Once you combine this technology with machine learning, the users won’t have to deal with it. As a result, payments of bills will never be late, or get to the stage of a bank overdraft.”
Cook explains that he sees Intuit in the middle of all this. He gives an example of how up until now, the company focused on serving private people and small businesses by helping them solve bookkeeping problems. In recent years, however, they noticed a growing group of freelancers – 750 million worldwide, according to Cook – who must monitor their income and expenses carefully. To address this sector, Intuit recently launched a new system, called QuickBooks Self-Employed, to simplify the process for them. For example, the system enables self-employed people to organize their receipts and monitor them using smartphone photographs. The application stores the photos and feeds in the particulars of the transaction, so that they can be retrieved easily in the tax season. Another feature tracks how many kilometers self-employed people drive in a car for work purposes. Intuit says this is likely to save over $7,000 a year per self-employed person.
The future, Cook posits, lies in taking the difficult problems – the bureaucracy, the cumbersomeness that no one wants to deal with – and turning them into something visual, simple, and intuitive for the user, and in as automated a way as possible.
Incidentally, if you are wondering about the availability of the products in Israel, we are sorry to disappoint you, but it is not really on the agenda.
After 6 years, 86 Million Dollars in funding, a peak of 23 Million users and 4 different failed products, today marked another chapter for Mobli, with the sale of one of its patents to Snap Inc.
Mobli, which back in the day was considered one of Instagram’s leading competitors shut down mid-last year after its last product (Galaxia) failed.
According to multiple sources, Serial entrepreneur and investor Moshe Hogeg, who co-founded Mobli, signed the deal earlier this month, transferring Mobli’s Geofilters patent over to Snapchat for a sum of 7.7 Million dollars.
The patent (US#20160373805, US#9459778) was filed back in 2012 and was named Methods and Systems of providing visual content editing functions. To be more specific, the patent focused on what is known today as “Geofilters”, special photo filters that are available by mobile geo-location.
Why are Geofilters so important for Snapchat?
Besides being one of the popular features within the app, according to Snap Inc. recent S-1 Filing, Sponsored Geofilters are an integral part of the company’s revenues. As an example, Snap recognizes sponsored campaigns by Starbucks and Wendi’s, that made use of the Sponsored Geofilters and have created value for the advertiser.
“Starbucks also created an in-store experience throughout the campaign, using regularly-updated Sponsored Chain Geofilters available at their locations across the U.S. to promote the blended version of the drink. The summertime Sponsored Geofilters were a testament to the power of Creative Tools that let users share what they’re up to with their friends and family, with over 40 million people viewing Snaps that featured one of the Sponsored Geofilters.”
“Wendy’s blanketed its U.S. stores with Sponsored Geofilters that promoted the Jalapeño Fresco Chicken Sandwich.”
Since Mobli’s original patent for Geofilters was approved back in 2014, which is way before Snapchat planned the Geofilters feature, the company was actually infringing Mobli’s patent and we can only assume that the IP acquisition was done in order to avoid a potential infringement law-suit in the future.
Moshe Hogeg has declined to comment and we’re still waiting to hear from Snapchat. We’ll update this post as soon as we hear back.
The space startup ecosystem is blossoming, led by the wild success of Elon Musk’s SpaceX. The US is leading the charge, but startups from Israel, Singapore, the United Kingdom, and Finland are also putting the right foot forward to get “astropreneurship” off the ground. Despite their small size by comparison to the competition, Finland has made some stunning contributions to the industry in the last year, and one software company has entered the industry with several major goals in mind. Among them? Launching humanity’s first network of satellites around the Moon.
“Constellation size depends on use case: what kind of coverage or data acquisition rate is required,” Reaktor’s Juha-Matti Liukkonen, the company’s director of space and robotics, explained to Geektime. He oversees the company’s space spinoff, the aptly named Reaktor Space Lab (RSL), which is run by CEO Tuomas Tikka and COO Heikki Salokanto.
We first got the chance to chat last December on the sidelines of Helsinki’s annual Slush tech conference. In the haste to cover everything, we only managed to find the time recently to finish our initial exchange via email. That event showed off much of Finland’s space prowess, but could not do the local industry enough justice.
“European investment is seen mainly in Luxembourg, focusing on asteroid mining, but the European investment scene is starting to wake up to space as an opportunity,” Liukkonen comments, and Finland is an outsized contender in the burgeoning sector. “Finland has had a small and specialized space industry for over 30 years, mostly building instruments for ESA and NASA missions.”
RSL previously designed the Aalto-1 and Aalto-2 satellites for the local Aalto University. This July, they plan to launch Finland’s first fully commercial satellite — Reaktor Hello World — with ISRO from an Indian PSLV rocket. Liukkonen expects radio frequency permits for the satellite to clear the month before.
Their near-term business goal is to design and build satellite networks providing large amounts of data for projects like crop monitoring and smart cities. They offer the whole kit and caboodle: mission design, satellite design, manufacturing, testing, launch, and operations.
Liukkonen wouldn’t disclose the names of any prospective clients, but did say Space Labs has started designing basic network packages.
“We are designing one which seems to require between 20 and 28 satellites, and another one which can be operational at 10 satellites but whose capability and supported use cases will increase when more satellites get added.”
The instruments and abilities of those satellites would have to be sophisticated to maintain market share in a booming industry. Liukkonen thinks they are set on that front. “Hyperspectral data are key for many use cases from precision agriculture to mineral prospecting, and RSL can provide high spectral resolution and 10-30m GSD imagers as well.”
Deploying humanity’s first lunar satellite network
By spring 2018, we might be entering a new phase in the modern, entrepreneur-led space race. A handful of teams will compete for the Google Lunar XPRIZE in late 2017 and early 2018, trying to land a rover on the moon and get it to travel 500 meters across the surface.
Competitors will inevitably face a challenge from none other than Amazon founder Jeff Bezos’s Blue Origin and possibly (probably?) SpaceX. With companies like that making a push, the Moon is almost certain to start eating into the share of venture capital flowing to astropreneurial startups.
To that end, the ESA has put out calls for several lunar projects. One imagines the construction of a lunar base built exclusively with 3D-printing tech, while another would modify earth-observation (EO) technology by creating better imaging the moon (a description of any more detail would probably not qualify as ‘laymen’s terms’). A third, Lunar Cubesats for Exploration (LUCE), has Reaktor’s attention.
That project wants “to generate a number of cubesat / smallsat mission and system concepts . . . which can support ESA’s lunar exploration objectives and respond to flight opportunities in a timely way.” Lunar surface sensing, volatile prospecting, communications, and localization are all mentioned as desired satellite capabilities in the tender.
They also aren’t the only players in the game. NASA announced last week it would award two proposals for lunar cubesat concepts with resources to develop fully fledged mission schematics: “CubeX” by Suzanne Romaine of the Smithsonian Astrophysical Observatory would map elemental composition of the lunar surface while “BOLAS” by Timothy Stubbs of the NASA Goddard Space Flight Center would study the lunar hydrogen cycle and electromagnetic storms.
Both of those projects would deploy a 12-strong lunar satellite constellation. That means there are at least three complex proposals for building such a network right now, with several more probably being drawn up as you read this.
“Combining two cases that we are working on translates perfectly to a lunar application to support lunar operations,” Liukkonen conferred to Geektime.
Despite that and efforts by companies like would-be lunar-mining company and XPRIZE participant Moon Express, he does not see ‘lunarpreneurship’ happening without major pushes from governments.
The role of NASA and ESA in growing the lunar startup ecosystem
Various satellites have been put into lunar orbit historically, but only four are currently active: NASA’s Lunar Reconnaissance Orbiter, China’s Chang’e 5-T1, ARTEMIS P1 and ARTEMIS P2. Getting more up there necessitates NASA’s and ESA’s continued involvement, Liukkonen argues.
“We don’t see an immediate business opportunity in the Moon so these kind of missions need to be space agency initiatives. The space agency projects are needed to establish the necessary infrastructure on the Moon to enable later commercial operations.”
What kind of missions? Before anything like moon mining becomes sustainable, using the Moon as an interplanetary gas station is the most practical way to give industry a firm footing on Big Gray.
“In our view, one of the primary business opportunities in the Moon is to produce rocket fuel for the US, Chinese, and Indian Mars missions. Once we have an infrastructure to fuel spaceships in orbit, the economics of interplanetary flight changes and the solar system opens up.”
“Another interesting long term track is in-orbit or lunar spacecraft production so that we can avoid the expensive launch phase and related structural requirements for satellites,” he suggests. That means avoiding the excessive fuel costs of blasting out of Earth’s atmosphere, instead mechanizing ship production (or at least small-sat production) on the Moon or from a space station.
“This of course implies that we need to find and harvest most of the necessary materials from the Moon. Hyperspectral imagers help, but we’ll also need significant advances in robotics and 3D printing,” he says, dovetailing with the aforementioned feelers put out by ESA and a similar contest developed by NASA (watch the video) that gave out prizes last year and has a second iteration in 2017.
With the Lunar XPRIZE hanging in the air (in orbit?) and never-ending discussion about a mission to Luna as a stepping stone to Mars, the timeline for developing infrastructure is actually shorter than you might imagine. Building any sort of facilities will require ‘LO’ — lunar-observation satellite data — to inspect landing sites, possible base sites, and any area of the lunar surface that might be of interest for material exploitation in 3D construction.
“ESA wants to have humans back on the Moon by 2030,” Liukkonen. “That means we need to have the first rounds of infrastructure, such as radiation measurement satellites and other operations support systems appearing there around 2020.”
Coming in spring 2018: the ‘New Space’ race’s second wave
It might be better to separate space ventures over the last few years from what might happen from now on through the end of the decade. SpaceX has led the way in deploying reusable rockets and Planet (formerly Planet Labs) runs the gauntlet on large constellations of nanosats. But new players will definitely emerge as astropreneurs reach beyond LEO, or low-earth orbit.
“There is in fact a significant round of consolidation going on right now, the first wave of New Space companies are being acquired mostly by Planet,” Liukkonen posits, referring to their acquisition of companies like Terra Bella from Google. Satellite companies with an emphasis on data have been the focus of investors and incubators like ESA BIC. “World wide, there are many space companies bubbling under and on surface too, starting from the obvious OneWeb and Planet and ranging to smaller companies such as AstroDigital and RSL.”
Going forward, especially with the culmination of the XPRIZE battle, projects tied to the Moon are certain to grab some share of venture capital devoted to space startups.
Some but by no means all moon-bound projects will be to build lunar satellites, but orbiters will be essential for data transmission, ground communication, and all the other possible use cases described above. Reaktor may very well win a contract from ESA. In the interim, true space enthusiasts and venture capitalists should maintain close eyes on the race back to the Moon for the moonshot that might be the next SpaceX.
There are plenty of languages Google has not yet added to its machine translation program, but they tend to be on the smaller end. Google Translate now has 103 languages in its repertoire and very few programs can possibly hope to compete with the giant. Microsoft Translator and Yandex.Translate have managed to add Native American and Russian languages, respectively, that Google has not. There are also others from India, China, and South Africa that Google has not gotten to but other players – big and small – have.
Here are ten languages Google still hasn’t added to Google Translate and where you can find them.
Believe it or not Google has missed the main language of Myanmar, but one company has not: iTranslate. They don’t have the notoriety, but they purport to have 60 million downloads to their name. May of those might be in the Apple Store, as Google Play only records 5 million. The company is not affiliated with Apple, though they seem to have no issue using an endorsement from Apple COO Jeff Williams on their homepage.
Burmese has about 33 million speakers, a fairly large population but hardly close to the world’s top rankings.
Despite an impressive effort, this non-speaker noticed that the vocabulary in the iTranslate Burmese memory bank is a bit lacking in English (the Hebrew example is far simpler).
2-4. Papiamento, Udmurt, Muri
The Russian answer to Google, Yandex is a search and internet giant in Moscow. They too have spent time developing their own open translation tool with a focus on deep learning techniques and technology that resembles Google Translate’s new neural-networking-based zero-shot translation.
That ‘Ja man!’ accent of Jamaican English is unique among dialects, but some people point out it can be downright impossible to understand for some English speakers. Once you reach the point where two dialects within a given language are “mutually unintelligible” (as in speakers of either dialect cannot understand the other), then you are actually dealing with a wholly separate language. That’s the case with Patois (Patwah), or what some linguists call Jamaican Creole. It’s hardly a standardized language, but the prolifiicty of online conversations gives Patwah speakers plenty of opportunity to standardize spelling.
Also known as Tswana, this southern African language has nearly 5 million speakers between Botswana and South Africa. TheWebValue, or TWV, has a Setswana-English translator app for Android, but the ratings are not exactly something to be proud about (1.8 out of 29 votes). Another app called Thanodi by Intellegere Media fares better with a 4-star rating but hasn’t been updated since August 2013. This is an under-served language in need of some serious machine translation attention.
7-10. Wu (including Shanghainese), Min Chinese, Gan Chinese, and Cantonese
Learning and translating “Chinese” is in high demand with enterprise companies, ostensibly to reach the untapped market of ‘1 billion’ people throughout China. Often times it’s lost in the shuffle that “Chinese” as most English speakers know it is actually better dubbed “Mandarin Chinese,” and it is far from the only true language spoken in China. Cantonese, the main language of Hong Kong, has tens of millions of speakers beyond the 8 million or so residents of the Asian financial hub.
Simply called “乡音” or “Dialect,” this Android-available app calls itself a “dialect exchange platform” (according to Google Translate, ironically) and will translate written and even spoken examples of Shanghainese (a form of Wu spoken in the Shanghai region), Cantonese, and Min Chinese. It also includes data on dialects rather than what are arguably independent languages like southwestern Mandarin (a.k.a. Sichuanese, in the Sichuan and Chongqing provinces) and Shenyang Mandarin (among other Northeastern varieties).
So-called “V2X” startup Autotalks bagged $30 million Series D funding round late last week in funding according to an announcement from the company. Their main physical product is an eponymously branded chipset that is built into vehicle models. They released a 2nd generation of the chipset in 2016, timely considering the explosion in connected cars and self-driving vehicles’ dominance of technology headlines worldwide.
“The selection process of chipsets for vehicles is a lengthy process in which we interact and work with both Tier-1s and car makers to embed our solution in new card models,” Autotalks CEO Hagai Zyss told Geektime by email. “The fact we provide a ‘pre-integrated’ solution makes the design process for car makers / Tier-1s a much easier task, allowing them to focus on their ‘added-value’ functionality.”
“We are very pleased with the completion of this financing round, supported by such strong syndicate. It demonstrates clear vote of confidence in Autotalks’ ability to execute and gain a leading market share,” Autotalks CEO Hagai Zyss wrote in a press release. “Our mission to equip vehicles with such lifesaving technology is now being adopted by the regulators and leading car manufacturers. I believe our chipsets will soon be part of most new vehicles worldwide.”
“V2X” (“vehicle-to-anything”) is just one acronym you’ll find connected to their work — “V2V” (“vehicle-to-vehicle”) and “V2I” (vehicle-to-infrastructure”) are a couple others. They’ve gotten their chips into things like the Audi AG roof antenna. They announced their autonomous-vehicle-specific product, the CRATON2, at the ITS World Congress in Melbourne last October. They have even gone so far as to integrate their chips into motorcycles.
While it is easier to integrate the chips into a vehicle when it’s built, they have gone out of their way to make sure they are adaptable to older models as well.
“These devices will allow older cars to be equipped with V2X capability – thus further enhancing the penetration of V2X to more vehicles,” said Zyss.
CB Insights wrote last July that automotive technology investments (including but not limited to connected cars) would probably break its annual investment benchmark by the time 2017 came around. V2V was considered a $15 billion market back in 2015 by Global Market Insights, with annual growth estimated at 5 percent per year through 2023. Their report explicitly names Autotalks as a major industry player along with other software companies like Qualcomm and Denso.
CB had tallied $450 million in deals into last summer, led by Zoox’s $200 million round from Lux Capital and DFJ. The only year that came close to this was 2013, when Mobileye raised $400 million (nearly two thirds of the industrial total four years ago).
Autotalks keeps pace with US regulations
They grabbed cash from some of Israel’s most prominent investors, including the Samsung Catalyst Fund, which is intensely focused on new car technologies these days. Alongside newcomer funders Fraser McCombs Venture and Vintage Investment Partners, the following Israeli high-rollers also added to money they have already given Autotalks: Magma Venture Capital, Gemini Israel Fund, Amiti Fund, Mitsui & Co. Global Investment, Liberty Media’s Liberty Israel Venture Fund, and Delek Motors.
Autotalks is not the first company to get money for this kind of tech by any means. Mobileye, acquired by Intel last week in a $15.3 billion deal constituting Israel’s largest tech acquisition ever, is one of the established players. Peloton Technology, Nauto, BestMile, Quanergy and Pittsburgh-based Cruise (acquired by General Motors) are among others.
One of the big safety uses for V2V is connecting 19-wheel freight trucks that can coordinate driving speeds or braking that might avoid high-impact highway crashes in the US. The US Department of Transportation (USDOT/DOT) has launched a V2V trial in Ann Arbor by Michigan University and the MIT Technology Review has called it one of 2015’s top breakthrough technologies.
Autotalks pointed out another recent USDOT policy that mandated DSRC — dedicated short-range communication — installation into all new light vehicles sold in the US by 2023. ” To meet this target,” the company says, “deployments will start in 2019.”
Small player versus Intel and Microsoft
Commoditization should position companies like Autotalks well for the near future, perhaps even encourage the company to move toward an IPO with a secure market position. With the latest investment being such a late stage round (Series D), the company should be plenty confident to go against the grain and leave the private markets. Big brother Mobileye made the move and obviously never lost its “acquit-ability,” if you will.
“In our view it just another proof of the value that exists in bringing innovative technologies into the auto industry,” Zyss said of the deal, about which he thinks plenty has been said. “We believe that our technology will play a key role in the accelerated move towards full autonomous driving. In parallel to the path to ‘full autonomy,’ V2X technology will help save many lives by addressing ‘non-line of site’/limited visibility use cases.”
While Intel is making a splash trying to corner car software, there is still plenty of competition. Toyota reached a deal to use the latter’s patents for future self-driving car development
“Microsoft invests $11.4 billion annually in research and development and for more than 30 years has been developing innovative technologies that are powering today’s connected car experiences,” asserted Erich Andersen, corporate VP and chief IP counsel of Microsoft’s Intellectual Property Group, said last week. “Microsoft doesn’t make cars; we are working closely with today’s car companies to help them meet customer demands…”
Every software company is jumping in right now, and limiting discussions about market dominance to Intel and Microsoft is probably unfair to say the least. As Mobileye has proven, a company that isn’t an American tech powerhouse can make an indisputable and impossible-to-ignore splash in an important technology market. With the experience and networks built by Autotalks for business, it’s plausible their actual networks could make them one of the startups to put on exit watch in the next few years.
Google is taking a massive logistical step to fix diversity gaps in Silicon Valley through a new partnership with Howard University, which will open a campus at the Googleplex.
Howard is one the US’s Historically Black Colleges and Universities (HBCUs) and will become a direct source of talent for future Google hires.
“Howard happens to be my alma mater, so I am especially proud to share that our formal recruiting from the university has evolved into a residency for Black CS majors right here at the Googleplex,” wrote Bonita Stewart, VP of Global Partnerships, on Google’s official blog. Being dubbed Howard West, select juniors and seniors from Howard’s computer science program will earn the right to spend up to three months at the new complex and receive instruction not only from Howard staff but also senior Google engineers.
“With a physical space on campus where Howard students and Googlers can grow together, I can only imagine what innovation and creativity will come to light,” Steward said, adding the development is built on years of cooperation through the Google in Residence (GIR) program that embeds company engineers at HBCUs throughout the US. “Through GIR we’ve learned a lot about the hurdles Black students face in acquiring full-time work in the tech industry. The lack of exposure, access to mentors and role models are critical gaps that Howard West will solve.”
Howard University President Dr. Wayne Frederick was equally enthusiastic and ambitious about the project.
“Howard West will produce hundreds of industry-ready Black computer science graduates, future leaders with the power to transform the global technology space into a stronger, more accurate reflection of the world around us,” Frederick wrote. “We envisioned this program with bold outcomes in mind—to advance a strategy that leverages Howard’s high quality faculty and Google’s expertise, while also rallying the tech industry and other thought leaders around the importance of diversity in business and the communities they serve.”
There are other programs directed specifically at historically black higher education, particularly the SXSW HBCU PArtnership led by Opportunity Hub Founder and CEO Rodney Thompson that brings about 500 HBCU students to Austin for the annual tech and startup confab. The Morehouse College Entrepreneurship Center offers the only HBCU-based certificate in the US related to startup investing (let alone at any school for that matter).
Howard was founded in 1867, right after the conclusion of the Civil War, and is the alma mater of American icons like Thurgood Marshall, Ta-Nehisi Coates, and Toni Morrison.
Despite the widespread belief there are not many African-Americans learning development or coding, HBCUs and other schools have graduated plenty of black computer science students. Roughly 20 percent of said grads came from HBCUs between 2001 and 2009 according to Bloomberg, but blakcs have composed only 1 percent of all technical employees at prominent Valley companies (including Google).
While racial bias might be cited as one issue, proximity is likely and unfairly another. San Jose State University, UC Berkeley, San Francisco State University, and Stanford were the top four most likely schools to produce Silicon Valley workers according to Jobvite. Riviera Partners ranks Berkeley, Stanford and UCLA in the top five for tech talent recruitment in the region. Outliers like MIT, Carnegie Mellon and the University of Waterloo are known for top tier robotics and AI research, leavbing many schools (including HBCUs) outside of recruiters’ itineraries.
This will help HBCU students get more exposure. It was not mentioned if Howard would expand its program beyond three months or if other schools would follow in their footsteps, but it is hard to imagine there won’t be an effort.
All in all, this seems like one of the most practical approaches to solving Silicon Valley’s diversity problem undertaken yet.
Google rolled out several updates specifically for Brazilian users at an even in São Paulo that include special Brazilian Portuguese translation of emoji searches, audio calling via Google Duo, and direct posting to Google Search by people and places in the US and Brazil. They also announced a $5 million grant to the Lemann Foundation for a technology education project in the country.
“Our engineering team in Belo Horizonte has made remarkable contributions to our products globally, such as improving health-related searches, wrote Google VP of Product Management Mario Queiroz on Google Blog. “But we know there is still a lot of work to do in Brazil and elsewhere to make technology work better for more people.”
Google maintains two offices in Brazil: Belo Horizonte and São Paulo. The announcements were made at the “Google for Brazil” event late Wednesday in the latter city. Brazil will be one of the first to get the audio-only version of Duo, a variation of Google Hangouts that more resembles FaceTime. But low-coverage areas make Duo unusable, so requests for an audio default have been answered by the technology giant.
“This feature will be available starting today first in Brazil, and we’ll be rolling it out to users around the world in the coming days,” Queiroz said. Another Hangouts variant, Allo, will also get a makeover in the South American superpower with easy file sharing for Android users (PDF, .docs, mp3s, APKs, and Zip files). Updates to the Smart Smiley feature will roll into speakers of Brazilian Portuguese, “which uses machine learning to help you find the right emoji faster. Tap the Smart Smiley icon on the compose bar, and the app will suggest relevant emojis and stickers to help you finish your thought.”
The direct-to-search posts are actually long overdue not only in Brazil but anywhere. Updates on a given website will appear in search results underneath the main URL result, in the same way direct links to an About page or Contact form might appear beneath a certain business site.
“So if you’re searching for the Henry Ford Museum in the U.S. or for Vanessa da Mata in Brazil, you’ll see updates directly from the source with relevant information, like new exhibits, timely updates and interesting facts. Beyond these categories in the U.S. and Brazil, we’ll continue to experiment globally and look forward to making Search even more useful and timely.”
The updates to Waze Brazil tie into Google’s recent global update to Maps: location sharing. The latter, which allows you to share your immediate location to contacts, should help with the introduction of Waze Carpool in Brazil later this year. Google last made country-specific updates for Brazilians last year with in-country travel information for Google Flights. Back in 2012, Google Maps added online touring for Brazilian pre-colonial cities across the country near present-day Fortaleza, Brasilia, Recife, Salvador, and Natal.
Google has updated its news and weather app to create a steady stream of 200 relevant headlines to individual users, the company announced on its blog late Tuesday. Dubbing it simply “More Headlines,” the idea is to more directly address readers’ interests by circumventing search and browsers.
“You’ll also enjoy fast-loading AMP articles, as an increasing number of publishers adopt the AMP format,” Google’s effort to make slow-loading or non-mobile websites more accessible to phone browsers. “As usual, each story retains the goodness of a comprehensive perspective—expand a card to gain insight from different articles such as Highly Cited, Local Source and Fact Check. Everything stays algorithmic—from clustering articles to classifying stories to ranking the stream.”
The move was announced just hours after a separate update to create search shortcuts on the Google app’s home screen, a move this author considered redundant. It could be these two moves, which in some ways contradict each other, is a way of A/B testing which more efficient method is preferred by users. But this particular move by Google to create a collated newspaper as it were, as simple and monotonous as it may sound, is actually an offensive tactic against their major competition in news dissemination: Facebook.
Google’s advantages versus Facebook in considering news authority
Facebook has gone out of its way to imply it has that crown by referring to its homepage as a newsfeed and constantly perfecting an algorithm that shows users content from websites they constantly visit. But Facebook’s management of that power has been, frankly, piss-poor. The social network has been on the receiving end of infinite criticism for allowing deliberately fake news websites to pass off as legitimate while also silo-ing readers from alternate sources or viewpoints.
It is not clear by any means Google is going to be revolutionary and try dropping a handful, dozen or so articles from less-than-preferred or not-so-often-visited news sources into individual news collections, but this might be a chance to do just that. Regardless of whether or not it does, Google is drawing on information from across the entire internet when it builds its feeds; Facebook is limited not only to its own network but also the patchwork of rarely overlapping connections respective of each individual Facebook account.
Google News is far from perfect. It often gives weight to well-known sources when deciding whose coverage to feature and which websites should be relegated to the ‘further reading’ section. Large, established news sources can simultaneously be considered reliable and yet vanilla, as up-and-coming or smaller websites will often take greater effort to add extra information or analysis into their articles in the (sometimes futile) hope Google will recognize its coverage of a particular event as even more authoritative (in the same sense Geektime will sometimes take more time to write its coverage of particular items to best more established technology news outlets like The Verge, Techcrunch, or Wired).
Regardless of those imperfections, Google has been arguably more receptive than Facebook to the challenge of downright false stories and sorting out authority of sources in general. This is despite assurances the social network would begin flagging disputed content and outright blocking illegitimate websites. But looked at in tandem to Facebook’s summer 2016 updates that de-emphasize news pages in favor of personal posts by friends and family, the common human errors of those peers promoting fake news they think is real is a danger Facebook still must grapple with.
Google is not without its sins, but has directly shut 200 URLs out of its ad network over fake news complaints. Google has also encouraged further ‘innovation’ in news publishing and reporting in Europe with its Digital News Initiative Innovation Fund:
“We’re looking for projects that demonstrate new thinking in the practice of digital journalism; that support the development of new business models, or maybe even change the way users consume digital news,” Google Blog explains. “Last round we issued a call for collaboration–across industry and across the region–and of course we’d love to see this trend continue.”
It’s not a new initiative, but it certainly fits into the trope many have taken (such as the New York Times and Washington Post) that journalism is more important than ever in the age of Donald Trump’s persistent dishonesty.
Facebook’s founder Mark Zuckerberg will have a tough time shaking off the ire of observers after his blanket claim in November 2016 fake news played no role in deciding the US presidential election.
Despite announced changes by Facebook either to de-emphasize news or to put the brakes on fake news sources, Facebook is eating time from users who could otherwise visit news websites during the day. Users are spending up to an hour a day on the social network according to Facebook’s Q1 2016 earnings report. His rush to play down the possibility is what frustrates people today, knowing how dominant Facebook is in deciding what viewers see and how they perceive it.
Other changes to Facebook’s like buttons that allow people to reflect on articles with an emoji also might draw more engagement. The more emotive the article headline, the more likely a reaction and focus from the reader. More emotionally gripping and ‘clickbaity’ headlines, sometimes sensational, were a major draw by fake news sources in 2016, and likely factored into Facebook’s decision to count poor headlines against content sites in deciding whether or not to give them weight.
Further, Zuckerberg’s recent attempt to play down the issue again seems to amplify his network’s credibility in tackling the problem.
“It’s not always clear what is fake and what isn’t,” he said last week. “A lot of what people are calling fake news are just opinions that people disagree with.”
Being unable to discern opinion from fact would put Facebook at a massive handicap against Google for reliability. If this is something troubling Facebook’s algorithm writers, readers will be more enticed to trust Google News updates instead.
Yes, Facebook and Google are the main arbiters of news on the web
Some might scoff at this view of Google and Facebook. How can a search engine and a social network be in direct competition over news? The two sites are the largest referrers of news traffic on the web according to a report by Parse.ly, hovering around 40 percent each in terms of traffic. Yahoo was in an abysmal third place at 5 percent.
That is unlikely to change in the foreseeable future, meaning any efforts to tweak news feeds by either site should be watched closely. Until Facebook introduces its own search engine and web browser combo, or Google finally gets its social network efforts right, it will be a push-and-pull duopoly by the two websites that controls how we get our new information about global events. Whether they want to put the effort in or not, the two companies have an immense responsibility to balance correct coverage with individual users’ favorite coverage.