Blog of the website «TechCrunch» Прогноз погоды

People

John Smith

John Smith, 49

Joined: 28 January 2014

Interests: No data

Jonnathan Coleman

Jonnathan Coleman, 32

Joined: 18 June 2014

About myself: You may say I'm a dreamer

Interests: Snowboarding, Cycling, Beer

Andrey II

Andrey II, 41

Joined: 08 January 2014

Interests: No data

David

David

Joined: 05 August 2014

Interests: No data

David Markham

David Markham, 65

Joined: 13 November 2014

Interests: No data

Michelle Li

Michelle Li, 41

Joined: 13 August 2014

Interests: No data

Max Almenas

Max Almenas, 53

Joined: 10 August 2014

Interests: No data

29Jan

29Jan, 32

Joined: 29 January 2014

Interests: No data

s82 s82

s82 s82, 26

Joined: 16 April 2014

Interests: No data

Wicca

Wicca, 37

Joined: 18 June 2014

Interests: No data

Phebe Paul

Phebe Paul, 27

Joined: 08 September 2014

Interests: No data

Артем Ступаков

Артем Ступаков, 93

Joined: 29 January 2014

About myself: Радуюсь жизни!

Interests: No data

sergei jkovlev

sergei jkovlev, 59

Joined: 03 November 2019

Interests: музыка, кино, автомобили

Алексей Гено

Алексей Гено, 8

Joined: 25 June 2015

About myself: Хай

Interests: Интерес1daasdfasf, http://apple.com

technetonlines

technetonlines

Joined: 24 January 2019

Interests: No data



Main article: Nature

<< Back Forward >>
Topics from 1 to 10 | in all: 48

European founders look to new markets, aim for profitability

09:47 | 12 February

To get a better sense of what lies ahead for the European startup ecosystem, we spoke to several investors and entrepreneurs in the region about their impressions and lessons learned from 2019, along with their predictions for 2020.

We asked for blunt responses — and we weren’t disappointed.

These responses have been edited for clarity and length.


Kenny Ewan, founder/CEO, Wefarm (London)

I’ve often been faced with questions around how we can generate revenue in markets like Africa. There has historically been a view that you can do something good, or you can generate revenue — and companies that talk about developing markets usually get squarely lumped into the former. While mission-led companies achieving tremendous growth has been talked about for a while, 2019 has been a year I have felt conversations with investors and others really begin to shift to the reality of that and it’s thanks to more and more proof points being delivered by startups across the board.

As more and more businesses begin to realize they don’t need to wait for the internet to descend from the sky for these markets to become hubs of commerce and innovation — and see that it’s already happening — I believe 2020 will continue to witness more and more historic tech companies shifting their focus to markets like Africa and that there will be more coverage and discussion as a result.

 


0

Compound’s Mike Dempsey on virtual influencers and AI characters

19:27 | 18 January

In films, TV shows and books — and even in video games where characters are designed to respond to user behavior — we don’t perceive characters as beings with whom we can establish two-way relationships. But that’s poised to change, at least in some use cases.

Interactive characters — fictional, virtual personas capable of personalized interactions — are defining new territory in entertainment. In my guide to the concept of “virtual beings,” I outlined two categories of these characters:

  • virtual influencers: fictional characters with real-world social media accounts who build and engage with a mass following of fans.
  • virtual companions: AIs oriented toward one-to-one relationships, much like the tech depicted in the films “Her” and “Ex Machina.” They are personalized enough to engage us in entertaining discussions and respond to our behavior (in the physical world or within games) like a human would.

Part 1 of 3: the investor perspective

In a series of three interviews, I’m exploring the startup opportunities in both of these spaces in greater depth. First, Michael Dempsey, a partner at VC firm Compound who has blogged extensively about digital characters, avatars and animation, offers his perspective as an investor hunting for startup opportunities within these spaces.

 


0

Google and ClimaCell team up to launch a new high-resolution weather forecast for India

19:00 | 7 January

Weather technology company ClimaCell, which is taking a number of innovative approaches for gathering weather data and building forecasting models, today announced that it is partnering with Google Cloud to launch a new high-resolution forecasting model. The first model will focus on India, with other geographies following soon. ClimaCell and Google will make the forecast available for free through under the Google Cloud Public Datasets program.

The model will provide forecasts for the next 48 hours and have a resolution of 2km and 15-minute timesteps. It will also serve as the foundation for other weather products from ClimaCell for predicting floods, air quality and more.

“For the first time in history, a private company is offering a full-blown operational numerical weather prediction model for an entire country, working continuously and providing high-resolution forecasts for up to 48 hours ahead,” said ClimaCell CEO and co-founder Shimon Elkabetz. “Not only is it an historical milestone, we are providing it completely free of charge. We invite others to join us in making weather data free and accessible for everyone.”

The company argues that it takes a local model like this to address the nuances of local climates and geographies. The company plans to customize the model for other regions over time.

Since the models are available as public datasets, anybody will be able to make use of them. What we don’t know yet, of course, is how good these forecasts are, but ClimaCell has developed a pretty good reputation over the last year or so and a number of large companies, including airlines like Delta, Jetblue and United. The company also offers a freemium mobile app for consumers.

 


0

Pachama launches to support global reforestation through carbon markets

00:20 | 4 January

The world’s forests are ablaze, under threat from illegal logging, and disappearing due to the less dramatic environmental degradation wrought by drought and other signs of climate change.

It’s part of the negative feedback loop that seems to be accelerating climate change as greenhouse gases accumulate in the atmosphere, but one startup company is trying to facilitate reforestation by supporting carbon offsets that specifically target the world’s flora.

Pachama has raised $4.1 million to create a marketplace where companies can support carbon offset projects. The company is backed by some big names in tech investment like former Uber executive Ryan Graves, through his private investment firm, Saltwater, and Chris Sacca, a prominent early investor in Uber, through his Lowercase Capital firm.

Founded by Diego Saez Gil, a serial entrepreneur whose last company was a startup selling a “smart-suitcase”, Pachama is aiming to bring reforestation projects to the carbon markets whose impacts can be independently verified by the company’s monitoring software to ensure their ability to offset emissions.

“We were making a smart connected suitcase which got banned,” says Saez-Gil. “After that I decided to take some time off and I was quite burnt out. I wanted to do some soul searching and tried to decide what i wanted to put my efforts.” 

He traveled to South America and did a trip through Amazon rainforest in Peru. It was there that Saez-Gil saw the effects of deforestation in an area that represents a huge carbon dioxide offset for the planet.

“There are about 1 billion hectares on the planet that could be reforested,” says Saez Gil.

That opportunity — to contribute to the perpetuation of independently validated carbon markets around the world is what convinced investors like Paul Graham, Justin Kan and Daniel Kan, Gustaf Alströmer, Peter Reinhardt, Jason Jacobs, Chris Sacca from Lowercase Capital, as well as funds such as Social+Capital, Global Founders Capital, and Atomico to contribute to the company’s $4.1 million funding.

It’s a pretty big consortium to finance what amounts to a small capital commitment (given the size of the funds under management that these investors have at their disposal), but investors are right to be a little wary.

Carbon markets are driven by policy and policymakers have been reluctant to draft legislation that would put a high enough price on carbon emissions to make those markets viable.

“Pachama’s carbon credit marketplace is launching at a pivotal moment when awareness of the climate crisis is reaching an all-time high, and businesses are increasingly looking to become carbon neutral,” said Ryan Graves, Pachama’s lead investor and new director said in a statement. “What attracted me to Pachama was the company’s use of technology to bring trust to an industry that desperately needs it, and gives the verifiable results to the purchasers of carbon credits.”

Awareness doesn’t equal political action however, and Pachama needs the political will of both governments and consumers to move the needle on creating viable carbon trading markets.

Pachama’s business becomes profitable only when the price of carbon moves beyond $15 per-ton of carbon dioxide (or similar emissions) offset. Currently, there are only two markets in the world where that threshold has been reached — the California market and Europe, according to Saez-Gil.

For Pachama’s founder, forest preservation and reforestation projects can have outsized benefits. “There are only 500 forest projects that are certified today… we need tens of thousands,” says Saez-Gil. “There are one billion hectares on the planet is available for reforestation without competing with agriculture.”

The restoration of native forests can contribute to replenishing global biodiversity and captures more carbon than cultivating forests for industrial use, but both are better than destruction to grow row crops or support animal husbandry, Saez-Gil says.  

Pachama sources projects that are approved by existing certification bodies, but offers its customers monitoring and management services through access to satellite imagery and sensors that provide information on emissions and carbon capture on reforested land.

It’s a potential solution to the problem of deforestation that’s plaguing countries like Brazil. “The government in Brazil they want to generate income for the country,” says Saez-Gil. If carbon markets paid as much as ranching, it would reduce the need for animal husbandry and plantation farming in Brazil, Indonesia, or place like Peru. 

Today, most investments in reforestation projects are done through middlemen, which increases opacity and the chance that projects are being double-counted or sold, according to Saez-Gil. Pachama has a person who is contacting forest project developers so that they can list the projects independently. Then the company verifies the offsets with satellite imaging systems.

The company currently has 23 forest projects — three in the Amazon rainforest in Brazil and Peru and projects in the U.S. in California, Vermont, New Jersey, Connecticut and Maine .

Saez-Gil has high hopes for the future of carbon markets based on demand coming, in part, from new regulations like those imposed on the airline industry.

“Airlines will have to offset part of their emissions as part of CORSIA,”  says Saez-il. That’s an offset of 160 million tons of emission per year. “There is all this demand coming for different offsets for different  markets that will make the price go up.”

 


0

Femtech startup Inne takes the wraps off a hormone tracker and $8.8M in funding

10:00 | 31 October

Berlin-based femtech startup Inne is coming out of stealth to announce an €8 million (~$8.8M) Series A and give the first glimpse of a hormone-tracking subscription product for fertility-tracking and natural contraception that’s slated for launch in Q1 next year.

The Series A is led by led by Blossom Capital, with early Inne backer Monkfish Equity also participating, along with a number of angel investors — including Taavet Hinrikus, co-founder of TransferWise; Tom Stafford, managing partner at DST; and Trivago co-founder Rolf Schromgens.

Women’s health apps have been having a tech-fuelled moment in recent years, with the rise of a femtech category. There are now all sorts of apps for tracking periods and the menstrual cycle, such as Clue and Flo.

Some also try to predict which days a women is fertile and which they’re not — offering digital tools to help women track bodily signals if they’re following a natural family planning method of contraception, or indeed trying to conceive a baby.

Others — such as Natural Cycles — have gone further down that path, branding their approach “digital contraception” and claiming greater sophistication vs traditional natural family planning by applying learning algorithms to cycle data augmented with additional information (typically a daily body temperature measurement). Although there has also been some controversy around aggressive and even misleading marketing tactics targeting young women.

A multi-month investigation by the medical device regulator in Natural Cycles’ home market, instigated after a number of women fell pregnant while using its method, found rates of failure were in line with its small-print promises but concluded with the company agreeing to clarify the risk of the product failing.

At issue is that the notion of “digital contraception” may present as simple and effortless — arriving in handy app form, often boosted by a flotilla of seductive social media lifestyle ads. Yet the reality for the user is the opposite of effortless. Because in fact they are personally taking on all of the risk.

For these products to work the user needs a high level of dedication to stick at it, be consistent and pay close attention to key details in order to achieve the promised rate of protection.

Natural contraception is also what Inne is touting, dangling another enticing promise of hormone-free contraception — its website calls the product “a tool of radical self-knowledge” and claims it “protect[s]… from invasive contraceptive methods”. It’s twist is it’s not using temperature to track fertility; its focus is on hormone-tracking as a fertility measure.

Inne says it’s developed a saliva-based test to measure hormone levels, along with an in vitro diagnostic device (pictured above) that allows data to be extracted from the disposable tests at home and wirelessly logged in the companion app.

Founder Eirini Rapti describes the product as a “mini lab” — saying it’s small and portable enough to fit in a pocket. Her team has been doing the R&D on it since 2017, preferring, she says, to focus on getting the biochemistry right rather than shouting about launching the startup. (It took in seed funding prior to this round but isn’t disclosing how much.)

At this stage Inne has applied for and gained European certification as a medical device. Though it’s not yet been formally announced.

The first product, a natural contraception for adult women — billed as best suited for women aged 28-40, i.e. at a steady relationship time-of-life — will be launching in select European markets (starting in Scandinavia) next year, though initially as a closed beta style launch as they work on iterating the product based on user feedback.

“It basically has three parts,” Rapti says of the proposition. “It has a small reader… It has what we call a little mouth opening in the front. It always gives you a smile. That’s the hardware part of it, so it recognizes the intensity of your hormones. And then there’s a disposable saliva test. You basically collect your saliva by putting it in your mouth for 30 seconds. And then you insert it in the reader and then you go about your day.

“The reader is connected to your phone, either via BlueTooth or wifi, depending on where you are taking the test daily… It takes the reading and it sends it over to your phone. In your phone you can do a couple of things. First of all you look at your hormonal data and you look at how those change throughout the menstrual cycle. So you can see how they grow, how they fall. What that means about your ovulation or your overall female health — like we measure progesterone; that tells you a lot about your lining etc. And then you can also track your fluids… We teach you how to track them, how to understand what they mean.”

As well as a contraception use-case, the fertility tracking element naturally means it could also be used by women wanting to get pregnant. Eirini Rapti

“This product is not a tracker. We’re not looking to gather your data and then tell you next month what you should be feeling — at all,” she adds. “It’s more designed to track your hormones and tell you look this is the most basic change that happens in your body and because of those changes you will feel certain things. So do you feel them or not — and if you don’t, what does it mean? Or if you do what does it mean?

“It builds your own hormonal baseline — so you start measuring your hormones and we go okay so this is your baseline and now let’s look at things that go out of your baseline. And what do they mean?”

Of course the key question is how accurate is a saliva-based test for hormones as a method for predicting fertility? On this Rapti says Inne isn’t ready to share data about the product’s efficacy — but claims it will be publishing details of the various studies it conducted as part of the CE marking process in the next few weeks.

“A couple more weeks and all the hardcore numbers will be out there,” she says.

In terms of how it works in general the hormone measurement is “a combination of a biochemical reaction and the read out of it”, as she puts it — with the test itself being pure chemistry but algorithms then being applied to interpret the hormonal reading, looping in other signals such as the user’s cycle length, age and the time of day of the test.

She claims the biochemical hormone test the product relies on as its baseline for predicting fertility is based on similar principles to standard pregnancy tests — such as those that involve peeing on a stick to get a binary ‘pregnant’ or ‘not pregnant’ result. “We are focused on specifically fertility hormones,” she says.

“Our device is a medical device. It’s CE-certified in Europe and to do that you have to do all kinds of verification and performance evaluation studies. They will be published pretty soon. I cannot tell you too much in detail but to develop something like that we had to do verification studies, performance evaluation studies, so all of that is done.”

While it developed and “validated” the approach in-house, Rapti notes that it also worked with a number of external diagnostic companies to “optimize” the test.

“The science behind it is pretty straightforward,” she adds. “Your hormones behave in a specific way — they go from a low to a high to a low again, and what you’re looking for is building that trend… What we are building is an individual curve per user. The starting and the ending point in terms of values can be different but it is the same across the cycle for one user.”

“When you enter a field like biochemistry as an outsider a lot of the academics will tell you about the incredible things you could do in the future. And there are plenty,” she adds. “But I think what has made a difference to us is we always had this manufacturability in mind. So if you ask me there’s plenty of ways you can detect hormones that are spectacular but need about ten years of development let alone being able to manufacture it at scale. So it was important to me to find a technology that would allow us to do it effectively, repeatedly but also manufacture it at a low cost — so not reinventing the whole wheel.”

Rapti says Inne is controlling for variability in the testing process by controlling when users take the measurement (although that’s clearly not directly within its control, even if it can send an in-app reminder); controlling how much saliva is extracted per test; and controlling how much of the sample is tested — saying “that’s all done mechanically; you don’t do that”.

“The beauty about hormones is they do not get influenced by lack of sleep, they do not get influenced by getting out of your bed — and this is the reason why I wanted to opt to actually measure them,” she adds, saying she came up with the idea for the product as a user of natural contraception searching for a better experience. (Rapti is not herself trained in medical or life sciences.)

“When I started the company I was using the temperature method [of natural contraception] and I thought it cannot be that I have to take this measurement from my bed otherwise my measurement’s invalid,” she adds.

However there are other types of usage restrictions Inne users will need to observe in order to avoid negatively affecting the hormonal measurements.

Firstly they must take the test in the same time window each time — either in the morning or the evening but sticking to one of those choices for good.

They also need to stick to daily testing for at least a full menstrual cycle. Plus there are certain days in the month when testing will always be essential, per Rapti, even as she suggests a “learning element” might allow for the odd missed test day later on, i.e. once enough data has been inputted.

Users also have to avoid drinking and eating for 30 minutes before taking the test. She further specifies this half hour pre-test restriction includes not having oral sex — “because that also affects the measurements”.

“There’s a few indications around it,” she concedes, adding: “The product is super easy to use but it is not for women who want to not think ever about contraception or their bodies. I believe that for these women the IUD would be the perfect solution because they never have to think about it. This product is for women who consciously do not want to take hormones and don’t want invasive devices — either because they’ve been in pain or they’re interested in being natural and not taking hormones.”

At this stage Inne hasn’t performed any comparative studies vs established contraception methods such as the pill. So unless or until it does users won’t be able to assess the relative risk of falling pregnant while using it against more tried and tested contraception methods.

Rapti says the plan is to run more clinical studies in the coming year, helped by the new funding. But these will be more focused on what additional insights can be extracted from the test to feed the product proposition — rather than on further efficacy (or any comparative) tests.

They’ve also started the process of applying for FDA certification to be able to enter the US market in future.

Beyond natural contraception and fertility tracking, Inne is thinking about wider applications for its approach to hormone tracking — such as providing women with information about the menopause, based on longer term tracking of their hormone levels. Or to help manage conditions such as endometriosis, which is one of the areas where it wants to do further research.

The intent is to be the opposite of binary, she suggests, by providing adult women with a versatile tool to help them get closer to and understand changes in their bodies for a range of individual needs and purposes.

“I want to shift the way people perceive our female bodies to be binary,” she adds. “Our bodies are not binary, they change around the month. So maybe this month you want to avoid getting pregnant and maybe next month you actually want to get pregnant. It’s the same body that you need to understand to help you do that.”

Commenting on the Series A in a supporting statement, Louise Samet, partner at Blossom Capital, said: “Inne has a winning combination of scientific validity plus usability that can enable women to better understand their bodies at all stages in their lives. What really impressed us is the team’s meticulous focus on design and easy-of-use together with the scientific validity and clear ambition to impact women all over the world.”

 


0

The founders of Robin Healthcare think doctors need smart assistants too

17:16 | 18 September

Robin Healthcare,  a new startup founded by serial entrepreneurs Noah Auerhahn and Emilio Galan, is hoping to harness the power of personal assistants to make the business of healthcare easier for the physicians who practice it

The company’s technology, which works much the same way as a Google Home or Amazon Alexa or Echo, is placed in hospital rooms and transcribes and formats doctor interactions with patients to reduce paperwork and streamline the behind-the-scenes part of the process that can drive doctors to the point of distraction, the company’s co-founder said.

“I had a background doing claims data work in healthcare was at UCSF and finishing my clinical training,” says Galan. “And I was hearing lots of doctors telling me not to practice.”

The problem, says Galan, was the overabundance of paperwork. After school Galan doubled down on his work in claims and billing launching a company called HonestHealth where he worked with institutions and companies . like The Robert Wood Johnson Foundation, Consumer Reports, and the New York State Department of Health to analyze health care claims data and develop consumer applications.

Galan met Auerhahn at the HLTH conference a few years ago just as Auerhahn was looking for his next challenge after the sale of his previous company, ExtraBux.

The two men saw the wave of smart devices coming and figured there must be a way to use the technology to build a fully billable clinical report from monitoring the conversations with patients.

The company currently has dozens of its smart devices installed in hospitals around the country including a large surgical practice in Tennessee, the Campbell Clinic; Duke University Medical Center’s Private Diagnostic Clinic; the University of California San Francisco Medical Center; and Webster Orthopedics in Northern California.

Robin integrates with the major electronic health records companies, Epic and Cerner, through third party integrations that are designed to make it easier to input data automatically as doctors are assessing a patient’s condition and delivering treatments.

Part of why Robin exists is to avoid technology interrupting care,” says Auerhahn. “Having fewer interactions with EMR is a good way to do that.”

Robin’s service is human-assisted natural language processing to make sure that the data is input correctly.

The company’s early vision has been enough to attract investors like Norwest Venture Partners, which led the company’s $11.5 million Series A round.

In all, Robin Healthcare has raised $15 million in financing. The company’s other investors include Social Leverage, the early stage investment firm founded by Howard Lindzon.

 

 


0

Energy Vault raises $110 million from SoftBank Vision Fund as energy storage grabs headlines

02:30 | 15 August

Imagine a moving tower made of huge cement bricks weighing 35 metric tons. The movement of these massive blocks is powered by wind or solar power plants and is a way to store the energy those plants generate. Software controls the movement of the blocks automatically, responding to changes in power availability across an electric grid to charge and discharge the power that’s being generated.

The development of this technology is the culmination of years of work at Idealab, the Pasadena, Calif.-based startup incubator, and Energy Vault, the company it spun out to commercialize the technology, has just raised $110 million from SoftBank Vision Fund to take its next steps in the world.

Energy storage remains one of the largest obstacles to the large-scale rollout of renewable energy technologies on utility grids, but utilities, development agencies and private companies are investing billions to bring new energy storage capabilities to market as the technology to store energy improves.

The investment in Energy Vault is just one indicator of the massive market that investors see coming as power companies spend billions on renewables and storage. As The Wall Street Journal reported over the weekend, ScottishPower, the U.K.-based utility, is committing to spending $7.2 billion on renewable energy, grid upgrades and storage technologies between 2018 and 2022.

Meanwhile, out in the wilds of Utah, the American subsidiary of Japan’s Mitsubishi Hitachi Power Systems is working on a joint venture that would create the world’s largest clean energy storage facility. That 1 gigawatt storage would go a long way toward providing renewable power to the Western U.S. power grid and is going to be based on compressed air energy storage, large flow batteries, solid oxide fuel cells and renewable hydrogen storage.

“For 20 years, we’ve been reducing carbon emissions of the U.S. power grid using natural gas in combination with renewable power to replace retiring coal-fired power generation. In California and other states in the western United States, which will soon have retired all of their coal-fired power generation, we need the next step in decarbonization. Mixing natural gas and storage, and eventually using 100% renewable storage, is that next step,” said Paul Browning, president and CEO of MHPS Americas.

Energy Vault’s technology could also be used in these kinds of remote locations, according to chief executive Robert Piconi.

Energy Vault’s storage technology certainly isn’t going to be ubiquitous in highly populated areas, but the company’s towers of blocks can work well in remote locations and have a lower cost than chemical storage options, Piconi said.

“What you’re seeing there on some of the battery side is the need in the market for a mobile solution that isn’t tied to topography,” Piconi said. “We obviously aren’t putting these systems in urban areas or the middle of cities.”

For areas that need larger-scale storage that’s a bit more flexible there are storage solutions like Tesla’s new Megapack.

The Megapack comes fully assembled — including battery modules, bi-directional inverters, a thermal management system, an AC breaker and controls — and can store up to 3 megawatt-hours of energy with a 1.5 megawatt inverter capacity.

The Energy Vault storage system is made for much, much larger storage capacity. Each tower can store between 20 and 80 megawatt hours at a cost of 6 cents per kilowatt hour (on a levelized cost basis), according to Piconi.

The first facility that Energy Vault is developing is a 35 megawatt-hour system in Northern Italy, and there are other undisclosed contracts with an undisclosed number of customers on four continents, according to the company.

One place where Piconi sees particular applicability for Energy Vault’s technology is around desalination plants in places like sub-Saharan Africa or desert areas.

Backing Energy Vault’s new storage technology are a clutch of investors, including Neotribe Ventures, Cemex Ventures, Idealab and SoftBank.

 


0

Google details AI work behind Project Euphonia’s more inclusive speech recognition

01:48 | 14 August

As part of new efforts towards accessibility, Google announced Project Euphonia at I/O in May: An attempt to make speech recognition capable of understanding people with non-standard speaking voices or impediments. The company has just published a post and its paper explaining some of the AI work enabling the new capability.

The problem is simple to observe: The speaking voices of those with motor impairments, such as those produced by degenerative diseases like amyotrophic lateral sclerosis (ALS), simply are not understood by existing natural language processing systems.

You can see it in action in the following video of Google research scientist Dimitri Kanevsky, who himself has impaired speech, attempting to interact with one of the company’s own products (and eventually doing so with the help of related work Parrotron):

The research team describes it as following:

ASR [automatic speech recognition] systems are most often trained from ‘typical’ speech, which means that underrepresented groups, such as those with speech impairments or heavy accents, don’t experience the same degree of utility.

…Current state-of-the-art ASR models can yield high word error rates (WER) for speakers with only a moderate speech impairment from ALS, effectively barring access to ASR reliant technologies.

It’s notable that they at least partly blame the training set. That’s one of those implicit biases we find in AI models that can lead to high error rates in other places, like facial recognition or even noticing that a person is present. While failing to include major groups like people with dark skin isn’t a mistake comparable in scale to building a system not inclusive of those with impacted speech, they can both be addressed by more inclusive source data.

For Google’s researchers, that meant collecting dozens of hours of spoken audio from people with ALS. As you might expect, each person is affected differently by their condition, so accommodating the effects of the disease is not the same process as accommodating, say, a merely uncommon accent.

A standard voice-recognition model was used as a baseline, then tweaked in a few experimental ways, training it on the new audio. This alone reduced word error rates drastically, and did so with relatively little change to the original model, meaning there’s less need for heavy computation when adjusting to a new voice.

The researchers found that the model, when it is still confused by a given phoneme (that’s an individual speech sound like an e or f), has two kinds of errors. First, there’s the fact that it doesn’t recognize the phoneme for what was intended, and thus not recognizing the word. And second, the model has to guess at what phoneme the speaker did intend, and might choose the wrong one in cases where two or more words sound roughly similar.

The second error in particular is one that can be handled intelligently. Perhaps you say “I’m going back inside the house,” and the system fails to recognize the “b” in back and the “h” in house; it’s not equally likely that you intended to say “I’m going tack inside the mouse.” The AI system may be able to use what it knows of human language — and of your own voice or the contest in which you’re speaking — to fill in the gaps intelligently.

But that’s left to future research. For now you can read the team’s work so far in the paper “Personalizing ASR for Dysarthric and Accented Speech with Limited Data,” due to be presented at the Interspeech conference in Austria next month.

 


0

Agolo attracts Microsoft and Google funding with AI-powered summarization tools

21:27 | 23 July

As the ways we consume content multiply and change, media creators are hard pressed to adapt their methods to take advantage. Short-form audio and video news is one growing but labor-intensive niche — and Agolo aims to help automate the process, pulling in the AP as a client and Microsoft, Google, and Tensility as investors.

Agolo is an AI startup focused on natural language processing, and specifically how to take a long article, like this one, and boil it down to its most important parts (assuming there are any). Summarization is the name of the process, as it is when you or I do it, and other bots and services do it as well. Agolo’s claim is to be able to summarize quickly and accurately, producing something of a quality worthy of broadcast or official documentation. Its deal with the AP provides an interesting example of how this works, and why it isn’t as simple as picking a few representative sentences.

AgoloThe AP is, of course, a huge news organization and a fast-moving one. But its stories, while spare as a rule, are rarely concise enough to be read aloud by a virtual assistant when its user asks “what’s the big news this morning?” As a result, AP editors and writers manually put together scores or hundreds of short versions of stories every day specifically for audio consumption and other short-form contexts.

Since this isn’t a situation where creative input is necessarily required, and it must be done quickly and systematically, it’s a good fit for an AI agent trained in natural language. Even so it isn’t as easy as it sounds, explained Agolo co-founder and CEO Sage Wohns.

“The way that we have things read to us is different from the way we read them. So the algorithm understanding that and reproducing it is important,” he said. And that’s without reckoning with the AP’s famous style guide.

“This is one of the most important points that we worked on with them,” Wohns said. “The AP has their style bible, and it’s a brick. We have a hybrid model that has algorithms pointed at each of those rules. We never want to change the language, but we can shorten the sentence.”

Agolo Listenable 1

That’s a risk with algorithmic summarizing, of course: that in “summarizing” a sentence you change its meaning. That’s incredibly important in the news, where the difference between a simple statement of fact and an egregious error can easily be in a single word or phrase. So the system is careful to preserve meaning if not necessarily the exact wording.

While the AP may not be given, as I am, to circumlocutions, it may still be beneficial to shift things a bit, though. Agolo worked closely with the news organization to figure out what’s acceptable and what’s not. A simple example would be changing something like “Statement,” said the source to The source said “Statement.” That doesn’t save any space, but you get the idea: essentially lossless compression of language.

If the AP team can trust the algorithm to produce a well-worded summary that follows their rules and only takes a quick polish by an editor, they could serve and even grow the demand for short-form content. “The goal is to enable them to create more content than was humanly possible before,” said Wohns.

The investment from and collaboration with Google falls along these lines as well, though not as laser-focused on turning news stories into sound bites.

“What we’re working on with them is making the web listenable,” said Wohns. “Right now you can ask Google a question but it often doesn’t have an answer it can read back to you.”

It’s primarily a bid to extend the company’s Assistant product as it continues its combat with Alexa and Siri, but may also have the extremely desirable side effect of making the data Google indexes more accessible to blind users.

The scope of Google’s data (Agolo is probably now getting the full firehose of Google News, among other things) means that the AI model being used has to be lightweight and quick. Even if it takes only ten seconds to summarize every article, that gets multiplied thousands of times in the complex workings of sorting and displaying news all over the world. So Agolo has been very focused on improving the performance of its models until they are able to turn things around very quickly and enable an essentially real-time summary service.

Agolo Research Application

This has a secondary application in large enterprises and companies with large backlogs of data like documentation and analysis. Microsoft is a good example of this: After decades of running an immense software and services empire, the number of support docs, studies, how-tos, and so on are likely choking its intranet and search may or may not be effective on such a corpus.

NLP-based agents are useful for summarizing, but part of that process is, in a way, understanding the content. So the agent should be able to produce a shorter version of something, but also tell you that it’s by this person (useful for attribution); it’s about this topic; it’s from this date range; it applies to these version numbers; its main findings are these; and so on and so forth.

Not all this information is useful in all cases, of course, but it sure is if you want to digest 30 years of internal documentation and be able to search and sort it efficiently. This is what Microsoft is using it for internally, and no doubt what it intends to apply it to as part of future product offerings or partnerships. (Semantic Scholar has applied a similar approach to journals and academic papers.)

It would also be helpful for, say, an investment bank analyst or other researcher, who can use Agolo’s timeline to assemble all the relevant documents in order, grouped by author or topic, with the salient information surfaced and glanceable. One pictures this as useful for Google News as well in browsing coverage of a specific event or developing story.

The new (undisclosed amount of) funding has Microsoft (M12 specifically) returning, with Google (Assistant Investment Group specifically) and Tensility Venture Partners joining for the first time. The cash will be used in the expected fashion of a growing startup: chasing sales and a few key hires.

“It’s about building out the go-to-market side, and the core NLP abilities of the team, specifically in New York and Cairo,” said Wohns. “Right now we’re about a 90 percent technical team, so we need to build out the sales side.”

Agolo’s service seems like a useful tool for many an application — anywhere you have to reduce a large amount of written content to a smaller amount. Certainly that’s common enough — but Agolo will need to prove that it can do so as non-destructively and accurately as it claims with a wide variety of datasets, and that this process contributes to the bottom line more than the time-tested method of hiring another intern or grad student to perform the drudgery.

 


0

World View’s sub-space high-altitude balloon lasts over two weeks aloft

18:55 | 5 June

World View is a company focused on alternative methods of doing some of the work traditionally handled by satellites – but just a bit closer to Earth. Its ‘Stratollites’ are balloons that can ascend to heights in Earth’s stratosphere (basically at the edge of space) and one key ingredient towards making them more commercially viable is ensuring they can spend a long time up in the air without coming down.

To that end, the company just completed its longest ever mission for a Stratollite continuously in flight: 16 days, according to The Verge, which is still a ways off from its goal of 60 days but much longer than its previous record of five days continuously aloft.

The Earth’s stratosphere starts at about 23,000 feet at the north and south poles, and at as high as 66,000 feet at the equator, and World View’s Stratollite operates between 50,000 and 75,000 in their ideal functional configuration. The plan from the company is to replace a lot of the work done by low-Earth orbit satellites, including high-altitude site photography, and eventually even to ferry passengers to those lofty heights.

World View’s latest trip included tests of its ability to stay relatively in-place despite variable high-altitude winds. It managed to stay in an area roughly 75 miles square for a period of eight days, and zeroed in on a much smaller target of just six miles wide for a total of 6.5 hours. These are crucial for its long-term success, since once of the values it provides over satellites is the ability to hover over a defined area for an extended period.

The company debuted its Tucson HQ and launch site in 2017, and a few months later flew a record flight of just 27 hours, so it’s come a long way in two years for a relatively new and unproven technology, which it hopes commercialize by 2020.

 


0
<< Back Forward >>
Topics from 1 to 10 | in all: 48

Site search


Last comments

Walmart retreats from its UK Asda business to hone its focus on competing with Amazon
Peter Short
Good luck
Peter Short

Evolve Foundation launches a $100 million fund to find startups working to relieve human suffering
Peter Short
Money will give hope
Peter Short

Boeing will build DARPA’s XS-1 experimental spaceplane
Peter Short
Great
Peter Short

Is a “robot tax” really an “innovation penalty”?
Peter Short
It need to be taxed also any organic substance ie food than is used as a calorie transfer needs tax…
Peter Short

Twitter Is Testing A Dedicated GIF Button On Mobile
Peter Short
Sounds great Facebook got a button a few years ago
Then it disappeared Twitter needs a bottom maybe…
Peter Short

Apple’s Next iPhone Rumored To Debut On September 9th
Peter Short
Looks like a nice cycle of a round year;)
Peter Short

AncestryDNA And Google’s Calico Team Up To Study Genetic Longevity
Peter Short
I'm still fascinated by DNA though I favour pure chemistry what could be
Offered is for future gen…
Peter Short

U.K. Push For Better Broadband For Startups
Verg Matthews
There has to an email option icon to send to the clowns in MTNL ... the govt of India's service pro…
Verg Matthews

CrunchWeek: Apple Makes Music, Oculus Aims For Mainstream, Twitter CEO Shakeup
Peter Short
Noted Google maybe grooming Twitter as a partner in Social Media but with whistle blowing coming to…
Peter Short

CrunchWeek: Apple Makes Music, Oculus Aims For Mainstream, Twitter CEO Shakeup
Peter Short
Noted Google maybe grooming Twitter as a partner in Social Media but with whistle blowing coming to…
Peter Short