2018 will see tech focus on making our devices not just smart, but AI smart
‘Deep learning will enter all fields,’ says Eyal Miller of Samsung NEXT Tel Aviv, as startups aim to work around the vast amount of processing power needed by algorithms
Shoshanna Solomon was The Times of Israel's Startups and Business reporter
In the coming year, artificial intelligence will make its way from centralized servers to our handheld devices and home gadgets, and will become a dominant force in all areas in which huge amounts of data are used, Israeli experts say.
AI, once a matter of data and algorithms fed to machines to make them think like a human, is used today for a wide range of applications, from facial recognition to detection of diseases in medical images to global competitions in games such as chess and Go.
And as the use of AI and machine learning grows, attention turns to making it cheaper, faster, more ubiquitous and available.
“Deep learning will enter all fields,” said Eyal Miller, the managing director and CEO of Samsung NEXT Tel Aviv, a local investment arm of the Korean giant Samsung Electronics that opened in Israel in 2016 and has invested in 10 Israeli startups, half of them in the field of AI.
“AI was already relevant in last few years of development in subcategories, such as natural language processing (NLP), computer vision (CV), and later in cognition, autonomous-ness and assistance-like interfaces. This trend will intensify and AI will be even more important and dominant across more, if not all, fields in 2018.”
Artificial intelligence is only just taking off as a field, but is growing at a compounded annual rate of almost 63 percent since 2016 and is expected to be a $16 billion market by 2022, according to MarketsandMarkets, a research firm.
Industries globally will have to adapt to computers taking over tasks traditionally done by humans, and the race is on for who will lead this technological transformation. Companies like Nvidia, Intel Corp., Samsung Electronics and Qualcomm Corp. are all competing for a piece of that lucrative space.
Israel is a considered a hotbed of talent, and is globally known for its abilities and research into artificial intelligence, high-performance computing and computer vision, both in academia and on the ground.
“Members of the Israeli academia and entrepreneurs understood the power of AI early on and took an important place on that stage,” said Shahar Tzafrir, a partner at TLV Partners, a venture capital firm that manages some $120 million in funds. He is constantly on the scout for breakthrough AI technologies to invest in, he said.
In addition, just as the Israeli army has been an incubator for software engineers and cybersecurity specialists, who honed their skills for military purposes and then exported them into the civilian world, it has done the same for developers of AI skills, which are now also spilling out into the Israeli tech scene, said Miller.
As the use of artificial intelligence explodes, the race is also on to create the chips on which the processors will run. US-based NVIDIA is the company whose chips currently dominate AI processing, a report by New York-based data company CB Insights said. But government-backed startups in China, US tech giants like Apple and Google, and existing semiconductor firms, like Intel Corp., are all eyeing the space. Developing chips to enable AI processors to run will be among the top trends the tech world will see in the coming year, CB Insights said.
And as these tech giants enter the field, they are making their AI platforms available to other developers and programmers.
“Many of the obstacles AI faced in its initial years — the massive processing power and electricity required to run the algorithms that enable computers to think — have been reduced,” said Miller. “Today almost any programmer can use AI in their work, just because giants like Google, Microsoft and Amazon are making their AI platforms available for whoever needs them.”
And as AI becomes accessible to all, companies are starting to integrate the technology into their products, and focus and effort are shifting to make our daily technology tools — like phones and cameras — not just smart, but AI smart.
To do that, though, these devices need to be able to power these fast-thinking algorithms.
Tel Aviv-based Brodmann 17 is an Israeli startup that is harnessing the new generation of AI development for our phones and other gadgets. The firm is developing neural-network based algorithms that will allow any device — from smartphones to smart cameras — to be powered by machine learning, thus bringing deep-learning based computer vision applications to the mainstream.
The technology was first developed at Tel Aviv University’s AI lab. The firm, founded in October 2016, has raised some $1.6 million to date, with some heavyweight investors backing it, including Samsung NEXT, Sony Innovations and Lool Ventures.
“AI and deep learning still consume a huge amount of processing power and electricity, and traditionally all of their applications have been dependent on the possibility of accessing servers, which have the capacity to power the algorithms necessary for processing the information,” said Adi Pinhas, the CEO of Brodmann 17.
“Our technology allows artificial intelligence algorithms to be run on edge devices, from robots to cellphones, without having to go via servers but with the same amount of accuracy in results,” he said. “And to do this we use just 5% of the power needed until now to work these algorithms.”
So, for example, instead of cameras sending loads of images from a road to a server which then processes the information to analyze traffic data, the cameras themselves will now be able to do this using Brodmann 17’s technology. “We transform edge devices into AI smart devices.”
The firm is already working with a number of firms to insert the technology into their devices, although Pinhas was mum about their identity.
Incidently, the name of the firm comes from that of German neurologist
Prof. Korbinian Brodmann who in 1909 defined the different areas in the human brain. Area 17, also known as Brodmann17, is the part of the brain responsible for processing visual information.
Other Israeli startups are harnessing the technology where it can help humans do their work quicker and better.
“All areas in which machines can replace or augment man, and that have sufficient datasets, will be affected by AI,” said Miller.
Tel Aviv-based Aidoc, for example, founded by graduates of the IDF’s elite Talpiot unit, uses machine learning to help radiologists better read X-rays. The technology analyzes all the patients’ relevant images and clinical data, allowing medics to get a more comprehensive picture, the company’s website says. Similarly, MedyMatch Medical Technologies Ltd. uses deep learning to help physicians better diagnose strokes. And then you have startups like Applitools, which uses artificial intelligence to help developers find flaws in the code they write.
“AI is a breakthrough technology that happens perhaps once in every 100 years,” Tzafrir said. “It will affect every part of our lives, from our jobs to medical diagnosis. We cannot exaggerate the importance of this revolution. AI will not anymore be something theoretical, but a necessity, working at the intersection between man and machines. AI won’t replace the human, but it will make them more productive, from identifying images in pictures to understanding texts or deciphering X-rays.”
