You may know Google I/O as ‘that yearly event where Google unveils a new Android version’, but it’s actually about much more. In fact, for the technology industry at large, the so called updates to Android are only a ‘development’ in the short term. A large number of people follow Google I/O to understand what the future of computing will look like. And this year, Google gave us a glimpse at that, with two announcements that will make a difference in the long term. This includes a new AI model called LaMDA and a step towards developing real commercial quantum computers.
What is LaMDA?
Let’s start with the development that will have meaning in the nearer term. LaMDA stands for Language Model for Dialogue Applications and is designed to improve the growing number of conversations we’re having with AI systems today. That includes things like the Google Assistant, or the myriad number of chatbots out there. In fact, whatever advancements Google makes here will also help platforms like Alexa and Siri in the long run.
“Language is remarkably nuanced and adaptable. It can be literal or figurative, flowery or plain, inventive or informational. That versatility makes language one of humanity’s greatest tools — and one of computer science’s most difficult puzzles,” the company said in a blog post.
With LaMDA, Google is claiming to bring the free flowing nature of human conversations to AI. Voice assistants like the Google Assistant may have improved over time, but they’re still stumped far too often. With LaMDA, Google is essentially trying to change the robotic conversations you have with a smart speaker to more free-flowing stuff.
“That meandering quality (of conversations) can quickly stump modern conversational agents (commonly known as chatbots), which tend to follow narrow, pre-defined paths. But LaMDA — short for “Language Model for Dialogue Applications” — can engage in a free-flowing way about a seemingly endless number of topics, an ability we think could unlock more natural ways of interacting with technology and entirely new categories of helpful applications,” the company claimed.
In the long term, it could reduce the learning curve associated with using smart speakers, voice assistants and chatbots. If systems like LaMDA, GPT-3 and BERT are perfected, they will essentially reduce the number of times voice assistants and chatbot come up with responses like “Umm, I didn’t get that” or respond with incoherent responses to a question.
LaMDA is ready now, but it will perhaps take another few years to truly evolve into a transformative system. But Google is looking at even bigger leaps in technology in the next 10 years. The company said it is going to build its first error corrected commercial quantum computer by the end of the decade. That’s a mighty short deadline for a technology that’s still pretty much in its nascent stages. For instance, Google said it wants to build a 1 million qubit computer, whereas its current systems sport less than 100 qubits.
Quantum computers are the next generation of computers, significantly faster than the fastest supercomputer we have today. While modern computers store information as zeros and ones (known as bits), quantum computers use concepts of quantum physics where particles can exist as both zero and one at the same time. These are known as qubits, and form the building block of quantum computers.
Google also unveiled its first Quantum AI campus, which is based in Santa Barbara, California in the US and will house the future of Google’s quantum research. “This campus includes our first quantum data center, our quantum hardware research laboratories, and our own quantum processor chip fabrication facilities. Here, our team is working to build an error-corrected quantum computer for the world,” the company said.
An error-correct quantum computer is a system that accounts for quantum errors through software. Qubits, explained above, are notoriously difficult to manage, and they lose their state very easily, which results in errors. Errors can also occur from environmental interference, electromagnetic fields and more. All of this is corrected using software, which is called quantum error correction, and makes quantum computers more dependable.
In the long run, a commercial quantum computer will allow Google to provide more services over the cloud. These may be enterprise services in the beginning, but given the significantly larger computational capabilities that these services have, they’ll make a significant difference to consumers too. It’s sort of how an advancement in core 5G technology will in the end improve how the Internet works on your phone.
Project Starline is more like LaMDA, in the sense that it’s ready and available now. However, it will be a while before Google sees consumer use cases for this.
The technology essentially builds a virtual chat booth, where users can appear as realistic 3D holograms instead of faces on computer screens. The technology uses an array of cameras and sensors to build these 3D models, so it’s not really something that can be done at home. It’s more like a modern take on old school photo booths, but could be set up in offices etc. It’s obviously a take on science fiction movies like Star Wars, where people appear as holograms when chatting with each other.
“Project Starline is currently available in just a few of our offices and it relies on custom-built hardware and highly specialized equipment,” the company said in a blog post. “We believe this is where person-to-person communication technology can and should go, and in time, our goal is to make this technology more affordable and accessible, including bringing some of these technical advancements into our suite of communication products,” it added.