How Microsoft, Google, AWS and Facebook are battling to democratise AI for developers
Source – computerworlduk.com
Developer conference season is in full flow, with Facebook’s F8 last week and Microsoft Build, Google I/O and Amazon Web Services holding a London Summit this week, and they all had one thing in common: the goal of making AI technology more accessible for developers on their platforms.
The big three cloud vendors made a slew of announcements this week, although AWS holds back most of its good stuff for re:Invent in Las Vegas later on in the year, all centring on machine learning and AI technologies.
The idea is that by simplifying complex and powerful AI technology like computer vision, natural language understanding and deep learning models, these tech giants can lock developers into their ecosystem and milk them for cash as they consume infrastructure and services.
So what did each of these companies say regarding AI this week?
Speaking in London yesterday, Amazon.com CTO Werner Vogels said: “Our mission is to make machine learning available and put it into the hands of every developer.”
The work AWS is doing here is wide ranging, offering developers everything from the computing power to run machine learning models on their infrastructure, all the way up to consumable services like Rekognition for image recognition, Polly for text-to-speech processing, and Lex for building virtual assistants and chatbots.
Then there is AWS SageMaker, announced at re:Invent last year with the aim of making it easier for customers to deploy machine learning algorithms.
It’s essentially a platform for authoring, training and deploying machine learning algorithms to business applications, without much of the manual heavy lifting generally involved, such as provisioning infrastructure and managing and tuning training models.
Google is a strong contender when it comes to AI and machine learning expertise, dating back to its open sourcing of the popular TensorFlow framework. It even recently rebranded its research division as Google AI.
Now, during Google I/O in Mountain View this week, CEO Sundar Pichai set his stall out clearly, stating that AI is enabling the company to pursue its core mission: “To make information more useful, accessible and beneficial to society.”
He followed this with a couple of striking consumer-facing applications of AI at Google, including a new feature for Gmail where it will complete whole emails, as well as the headline-grabbing Google Duplex, where Google Assistant can mimic humans by making phone calls, such as booking a table at a restaurant, for example.
Google is also focusing on the hardware side with its Tensorflow Processing Unit, or TPU, by announcing its new Cloud TPU.
At Build this year Microsoft CEO Satya Nadella continued to push his latest mantra: intelligent core, intelligent edge.
He outlined his goal to “commoditise AI” for Microsoft developers, saying: “What matters is can we translate these [breakthroughs] into frameworks, tools and services and put them into your hands as developers, so you can take AI and make an impact on every industry and application.
“That requires us to scale AI across both the cloud and the edge, we need the most productive toolchain to create and customise AI and you need openness for frameworks and infrastructure, there cannot be lock-in.”
Microsoft offers a similar stack to its rival AWS, from the Azure Machine Learning Studio for modelling, its wide range of ‘cognitive services’ for computer vision and text-to-speech, for example, to various deployment options for running these models on Azure infrastructure.
Where Nadella differs slightly is in his clear focus on IoT use cases and edge computing. Build was focused on new partnerships with Qualcomm to create a ‘vision AI developer kit’, including the hardware and software for developers to create camera-based IoT solutions, and drone company DJI to deliver complex compute at the device level.
Microsoft also announced Project Brainwave, a system for running AI models with specialised chips, allowing developers to deploy models on specialised silicon for higher performance than CPUs and GPUs. Microsoft says Brainwave puts Azure as the fastest cloud running real-time AI.
Facebook comes at this from a slightly different angle. It isn’t a cloud computing company but a social network, however it still has a vibrant developer community and a ton of R&D budget aimed at AI.
As outlined in a blog post during F8, Facebook sees AI as a “foundational technology”.
The company writes: “We’ve made deep investments in advancing the state of the art through scientist-directed research.”
Facebook also announced the next version of its open source AI framework: PyTorch 1.0.
There was one more consistent theme across the conferences this month: the importance of ethics in AI.
As the fallout from the Facebook data scandal continues to unravel, tech companies are keen to present themselves as trusted custodians of huge volumes of data, with these powerful technologies being layered on top.
Zuckerberg said: “We need to take a broader view of our responsibility, it’s not enough to just hold powerful tools, we need to ensure they are used for good, and we will.”
Google CEO Sundar Pichai picked up the same thread, saying: “We can’t just be wide eyed about the innovations technology creates. There are very real and important questions being raised about the impact of these advances. We know the path ahead needs to be navigated carefully and deliberately, and we feel a deep sense of responsibility to get this right.”
Satya Nadella’s comments were strikingly similar too, with the Microsoft CEO saying: “We also have the responsibility to ensure these technologies are empowering everyone and the technologies are creating equitable growth by ensuring that every industry is able to grow and create employment and also have a responsibility as a tech industry to build trust in technology.”
Under a slide titled ‘Ethical AI’ Nadella added: “We need to ask ourselves not only what computers can do but what computers should do. That time has come. We have formed an ethics board inside the company with a diverse group of people who govern the products we build and projects we engage in.
“We are investing to put tools in the hands of developers. Just like with good UI, we need good AI, to make this a first-class engineering discipline, where the choices we make can be good choices for our future.”
Microsoft also pledged $25 million to an “AI for accessibility” effort focused on how AI could help those with disabilities.
This is a good time of year to see where all of the tech giants (barring Apple for a few months yet) are focusing their attention. A few years ago it was all mobile, then it was voice interfaces and now the zeitgeist is firmly AI.
The interesting part if how tightly aligned these companies are in terms of where they see the future and how they can differentiate to attract developers and enterprises onto their platforms. Giving developers access to cutting edge technology in a consumable way is clearly seen as the next battleground for the technology industry, picking a winner at this stage would be a fool’s errand.
The quick dose of ethics on top can be seem more as a reflection of the fraught political moment we find ourselves in, and a response to the scaremongering AI that tends to be reflected in the mainstream, from job losses to Skynet.