All posts in “Developer”

Plasticity wants to help chatbots seem less robotic


Y Combinator backed Plasticity is tackling the problem of getting software systems to better understand text, using deep learning models trained to understand what they’re reading on Wikipedia articles — and offering an API for developers to enhance their own interfaces.

Specifically they’re offering two APIs for developers to build “more robust conversational interfaces”, as they put it — with the aim of becoming a “centralized solution” for Natural Language Processing (NLP). Their APIs are due to be switched from private to public beta on Monday.

“One thing where we think this is really useful for is conversational interfaces where you want to integrate real world knowledge,” says co-founder Alex Sands. “We think it’s also really useful when you want to provide instant answers in your application — whether that’s over the entire Internet or over a custom corpus.”

One example might be search engines that are competing with Google and don’t have their own instant answer technology. “[They] could use something like this. We’re in talks with a few of them,” notes Sands.

“The other application is conversational interfaces who want a new NLP stack that will give them a lot more information than what an academic package like Stanford CoreNLP would provide them today,” he adds.

A few years back, the founders worked on a hack project that expanded the powers of Apple’s AI voice assistant Siri, by adding support for custom commands — such as playing a Spotify track or dialing up the temperature via a Nest device. This was before Apple opened up Siri to third party apps so they were routing voice commands through a proxy — and claim to have basically built “the first app store for voice commands”.

The experience taught them that “NLP in general is not robust” for handling more complex commands and queries, says other co-founder Ajay Patel.

“The other problem was a lot of the natural language processing tools out there really take a simplistic approach to understanding what a user says,” he adds. “The most simplistic way to explain it is they’re looking for keywords to figure out what a user is asking.”

Plasticity is taking a different approach vs these keyword-based NLP systems; building a system that understands the semantics of a sentence so it can perform a linguistic breakdown — “to figure out all of the relationships and entities in a sentence”.

They can then hand that information to developers so they can build “more robust conversational interfaces around it”, as Patel puts it — such as, for example, a chatbot that’s more conversational and capable, given it can serve up answers it found online.

“Today you can ask Siri fact-based questions, like who directed a movie, or who a particular song. But you can’t ask it a more useful question, like when is Stanford Spring break?” he adds. “It can’t take a sentence from the Internet and then find the direct answer in that sentence and then return that to the user.”

Instead Siri usually performs a Google search and serves those results to the user — leaving users to do the last mile legwork of extracting an actual answer.

Plasticity’s promise is to cut out that last step by returning the right answer directly to the user.

“Our core technology uses deep learning to figure out the base level of NLP tags — so that’s things like parts of speech, syntax dependency tree. So we use machine learning on the base to figure that out, and we use TensorFlow and Google’s SyntaxNet module,” says Patel. “And then on top of that we’ve built custom C++ code that basically operates a lot more accurately and a lot faster than a lot of the competitors out there.”

Of course if the Internet is your oracle then there’s limitless scope to return not truthful answers but full-on falsities, fake news and other skewed and prejudiced views — as indeed we’ve already seen Google Home do. Oops. So how does Plasticity avoid its technology falling into a similar trap and ensure accuracy in the answers its API can help provide?

“What we do right now is we run it only over Wikipedia,” says Sands on this. “Then the plan from there is to slowly expand whilst still maintaining that accuracy that you’re talking about.”

The API has been more than 1.5 years in development at this point, and they claim “much higher accuracy and much higher speed” at parsing sentences than IBM Watson, for example.

Initially, Patel says they focused on areas that existing, keyword-based NLP systems we’re handling well — such as lists — and then continued building out the complexity to handle other “linguistic edge cases”.

While they name Google as their main competitor at this point — given the company’s stated aim of organizing the world’s information, building systems that can understand text is a clear necessity for Mountain View’s mission — even so they reckon there’s room for another NLP player to offer similar services to the wider market.

“[Google has] put a lot of work into understanding text on the Internet to do their instant answer question and answering… But we really think that there’s still a space in the market for a solution for everybody else out there, who’s not Google, who’s not putting in hundreds of millions of dollars of investment into machine learning — and we really think they’ve got no ambition to become a leader in NLP. For example Apple actually outsources their question and answering on Siri to Wolfram Alpha.

“So we think there’s a significant place in the market to be the natural language processing solution and knowledge graph solution for all the other artificial intelligence products out there.”

And while their first focus is on building NLP tech that can understand semantic structure and perform granular linguistic analysis, Patel says they may also expand to other areas — such as program synthesis — to add more abilities to the API in future.

Funding wise, they’re still in the process of closing out their seed but have taken funding from multiple investors at this point — including First Round Capital’s Dorm Room Fund and General Catalyst’s Rough Draft Ventures. They’ll be looking for more investment after YC demo day, they add.

Featured Image: Bryce Durbin

AssemblyAI wants to put customized speech recognition within reach of any developer


It’s clear that voice is becoming a major interface, as we witness the rise of the Amazon Echo, Google Home, Siri, Cortana and their ilk. We’re also seeing an increasing use of chat bots and other voice-driven tools, which often require speech recognition with a very specific vocabulary.

That’s where AssemblyAI, a member of the Summer ’17 Y Combinator class comes in. The startup is building an API that will help developers build customized chat interfaces quickly.

“We’re building an API for customized speech recognition. Developers use our API for transcribing phone calls or creating custom voice interfaces. We help them recognize an unlimited number of custom words without any training,” Dylan Fox, AssemblyAI’s founder told TechCrunch.

He says, most off-the-shelf speech recognition APIs are designed to be one size fits all. If you want to customize it, it gets really expensive. AssemblyAI hopes to change that.

When Fox was working at his previous job as an engineer at Cisco, he saw first-hand how difficult it was to create a speech recognition program with custom words. It usually involved a lot of engineering resources and took a long time. He came up with the idea of AssemblyAI as a way to make it easier, less costly and much faster. He added, that recent advancements in AI and machine learning have made it possible to do what his company is doing now.

It’s worth noting that the tool requires GPUs, rather than CPUs, for increased processing power because the task is so resource-intensive. Getting access to a sufficient number of GPUs to build and run the tasks has been a challenge for the three-person startup, but their affiliation with Y Combinator has helped in that regard. It’s also brand new tech, so they have to solve every problem they encounter on their own. There are no books to read or solutions to look up on Google.

Even though they are just three people, they believe user experience is going to be key to their success, so they have one team member fully devoted to developing the front end. They claim that no training is required to run the API. You just upload a list of terms or names and the API takes care of the rest.

Fox fully recognizes that it’s hard for startup to build a speech recognition tool without constantly worrying about the bigger companies swooping in and grabbing their market share, but he says his company is working hard to differentiate itself as a go-to tool for developers.

“As a smaller company focused on a speech recognition technology, we can provide a better experience [than the bigger companies].” He says that means paying attention to the little things that attract developers to a tool like better documentation, simpler integration and just making it easier to use overall.

So far the product is in private beta with several companies deploying it on GPUs in the cloud, but it’s early days. He says when the customers come, they will have to scale to meet those demands using additional cloud-based GPU resources. If it works as described, that shouldn’t be long now.

Featured Image: Bryce Durbin/TechCrunch

Qualcomm opens its mobile chip deep learning framework to all


Mobile chip maker Qualcomm wants to enable deep learning-based software development on all kinds of devices, which is why it created the Neural Processing Engine (NPE) for its Snapdragon-series mobile processors. The NPE software development kit is now available to all via the Qualcomm Developer Network, which marks the first public release of the SDK, and opens up a lot of potential for AI computing on a range of devices, including mobile phones, in-car platforms and more.

The purpose of the framework is to make possible UX implementations like style transfers and filters (basically what Snapchat and Facebook do with their mobile app cameras) with more accurate applications on user photos, as well as other functions better handled by deep learning algorithms, like scene detection, facial recognition, object tracking and avoidance, as well as natural language processing. Basically anything you’d normally route to powerful cloud servers for advanced process, but done locally on device instead.

Facebook is actually one of the developers that gained early access to the NPE and the social giant and is already seeing five-fold improvements in performance for its AR features on images and live video, when using Qualcomm’s Adreno GPUs on Snapdragon SoCs.

Qualcomm’s NPE works with the Snapdragon 600 and 800 series processor platforms, and supports a range of common deep learning frameworks including Tensorflow and Caffe2.

As more tech companies look for ways to shift AI-based computing functions from remote servers to local platforms in order to improve reliability and reduce requirements in terms of network connectivity, this could be a huge asset for Qualcomm, and a big help in maintaining relevance for whatever comes after mobile in terms of dominant tech trends.

Featured Image: Justin Sullivan/Getty Images

Google releases the final Android O developer preview


Google today launched the fourth and final developer preview of Android O, the latest version of its mobile operating system. As expected, there are no major changes in this release and, according to Google, the launch of Android O remains on track for later this summer. There’s still some time left before the official end of the summer (that’s September 22, in case you wondered), but given that Android Nougat was on a very similar schedule, I expect we’ll see a final release in August.

The final APIs for Android O arrived with the third preview release, so today’s update is all about incremental updates and stability. All of the major Android SDKs, tools and the Android Emulator will get minor version bumps in the next few days and the Android Support Library (version 26.0.0) is now considered stable, but, like before, the focus here is on making sure that developers can test their apps before the final version rolls out to users.

For users and developers, the new version of Android brings better notifications support across the OS, picture-in-picture support, autofill and more. There also are new features that are meant to optimize your phone’s battery. While none of the changes are revolutionary, Android developers should probably test their apps on Android O as soon as possible (even if they don’t plan to support the new features). To do so, they also should update to the latest version of Android Studio, Google’s IDE for writing Android apps.

The Google Play store is now also open for apps that are compiled against the latest API.

The Android O developer preview is available as an over-the-air update for regular users, too (assuming you are brave enough to run pre-release software on your phone). It’s available for Google’s Pixel, Pixel XL, Pixel C, Nexus 5X, Nexus 6P and the Nexus Player. To get it, you can enroll here.

Last year’s update, Android Nougat, now has around 11.5 percent market share in the Android ecosystem. It’s no secret that it takes the Android ecosystem quite a while to adapt new OS versions, but with a considerable number of Google’s own Pixel phones in the market now, it’s probably a good idea for developers to jump on the Android O bandwagon soon.

GM is putting app developers directly in the driver’s seat

GM is giving developers a fast lane to make it easier to build connected car apps for infotainment systems. 

The automaker is offering up its next-generation infotainment software development kit (NGI SDK) to the general development community with a new twist: App makers will actually be able to test their creations IRL with the new Dev Client program.

The automaker claims it’s the first time a car company is giving developers a shot to work on their apps in a real production vehicle this early in the process. The friendlier, more open platform could turn the GM dashboard into a new space for connected car innovation — if it catches on with developers.    

“By introducing GM Dev Client, we’re giving developers the missing link they need to finalize their applications,” said John McFarland, director of Global Digital Experience in a release.

The app creation process is streamlined, with an open developers network ready for new applicants. Once they’re ready to make something, developers can download the new SDK, which has been available since January, to build out their app and begin emulating the in-car environment to kick things off.   

Https%3a%2f%2fvdist.aws.mashable.com%2fcms%2f2017%2f7%2f47d979c5 25c6 4549%2fthumb%2f00001

Once an app design goes through GM’s internal review process, it can be downloaded to the developer’s own car for real-world testing. App makers will have to have at least one friend in the car with them, however, since safety features kick in so that a connected laptop can only be used in the passenger seat while a car is moving. 

GM is also planning to offer the SDK with a new set of templated frameworks, like a media player layout or a point of interest layout, to give developers a more focused starting point for projects. Those should roll out by the end of the year, according to the company.

Https%3a%2f%2fblueprint api production.s3.amazonaws.com%2fuploads%2fvideo uploaders%2fdistribution thumb%2fimage%2f80889%2f2946960b a019 4785 9493 7c012e807d39