One of the most hyped announcements at CES 2020 is Samsung Neon, a project with a simple goal: to create completely original digital (i.e., not actor-based) avatars that converse and learn until they're indistinguishable from humans. Such friendly and conversational AI would find much use in the hospitality and entertainment industries, though it can also be useful for any customer service role, from cashiers to baristas. But it's the simple but noble aspiration of Neon, the first company in semi-startup Samsung Labs STAR, that could set the project apart from previous AI robots. Its founder and president and CEO Pranav Mistry wants these creations to be friends of humanity. That is why Mistry refers to these AIs, called Neons, as a proper name, as new beings waiting to be refined in order to exist. He calls them artificial humans. "Neon is like a new kind of life," Mistry said in a press release. "There are millions of species on our planet, and we hope to add one more. Neon lights will be our friends, collaborators, and companions, continually learning, evolving, and forming memories of their interactions." This rampant futurism invites skepticism, and the response from Press since Neon's introduction hasn't been too kind. It's pretty clear that Neon's technology is in its infancy: while the STAR Labs booth at CES is packed with assorted Neon avatars, they're running through pre-defined routines and not ready to have freewheeling conversations. During a demo presentation, Mistry he asked one of his neon lights, who looked like a punk woman with a shaved head, a few questions that got adequate answers, but his facial expressions and mouth movements were awkward and unnatural. And yet, right next to it, another instance of the same Neon was going through his predefined routine with much more fluid movements. The comparison showed how far Neon has to go before people interact with its AI characters as the project plans. Because what is really important, what Neon has to take away, is the illusion. In order to converse with neon lights as we would with humans, we have to believe that we are conversing with something that can respond with enough context and fast speed.
How to make AI interact like humans
To Neon's credit, the team acknowledges that the project is in its infancy (Neon started only four months ago) and has a strategy for making its creations conversation partners. The first phase of Neon's timeline begins with the development of a core technology called Core R3, broken down into real-time, real-time, and responsive, which are archived guidelines that Neon follows. The reality directive means that talking to a neon should be like talking to a human, which includes limiting neon lights to human knowledge. They will be programmed not to be encyclopedias: no instant internet scan will be performed in the background. So if you ask an objective neon question, it will respond by asking if you want me to google it. Compared to all the smart assistants, this seems counterintuitive, the equivalent of having neon lights with one arm tied behind your back. But it's just the opposite of what Mistry and her team want them to have interactions with Neons. Think of it this way: you command Siri and Alexa; With Neon, Mistry wants you and an avatar to have conversations that develop memories and essentially some kind of proto-relationship. This doesn't mean neon lights are stupid, they'll just be loaded with knowledge modeled on the paper of each neon (at the CES booth, each neon had a name and title like ``student'' ; or ``stewardess & # 39; & # 39;). And for conversations to continue in real time (the second directive), this knowledge will be mostly local, thus avoiding the delay in getting your questions across to the cloud and vice versa. This should allow them to answer any question in 20 ms, an interval that humans can't really perceive. The other half of the credibility lies in the responses themselves: varied, unpredictable, natural. At least in the way they appear to us. The third directive, Responsive, means that the neon will gauge your emotions and vocal pitch to respond intuitively. They also have their own emotional spectrum, which Mistry and her team displayed on a graph during their presentation, and in their behind-the-curtain look, revealed a mind map of nodes representing the 7 million responses they contain. If you ask a neon the same question twice, it should answer differently. That's a lot of ambition, and far from what we saw capable neon lights at STAR Labs' CES booth. Towards the end of 2020, the group will augment its creations with SPECTRA, a technological group that comprises intelligence, learning, emotions, and memory, all of which should allow neon lights to grow into rounded digital individuals. Does this mean that neon lights will develop quirks? Personalities? The prejudices? It's too early to tell, and in truth, the team doesn't seem to know.
So... what will the neon lights be used for?
Neon's presentation at CES 2020 was an exploratory initiative: The team wanted the audience to respond and tell them what they thought neon lights could be used for, said Angie Chiang, technical editor and neuroscientist STAR Laboratories at LaComparacion. "In terms of applications, we need help. We are not targeting a specific area, so we need people who are experts in their field to help us make neon lights useful for them,” Chiang said. While installing Neons as customer service representatives seems like an obvious choice—essentially replacing chatbots with more user-friendly and customizable artificial humans—there's much more potential for deploying Neons in roles that could use more humanity when real humans aren't available. Chiang recalled his experiences volunteering at nursing homes, and installing a neon sign to keep the company sick during their recovery could be incredible palliative care. Neon lights could offer human interaction to people far away, the sick, the elderly, the lonely. They could fill in for human workers at inconvenient times, like late-night station anchors when the news comes in, Chiang suggests. They can be therapists or just confidantes (STAR Labs from the start ensures that confidentiality is built into the heart of Neon - only you and your Neon have access to your interactions, and they won't share your data without authorization). They could be actors, but instead of just producing neon lights and shipping them, STAR Labs will treat them as licensed individuals. So maybe a museum would secure a neon sign for an exhibit and choose to load it with relevant knowledge (you could even work as a museum guide). However, there are some concerns: first, that neon lights can take jobs. It is possible, of course, that companies are looking to cut costs by replacing customer support with online chatbots and automated help lines. There's also the question of how neon lights will affect job evaluation: will neon lights be cheaper than hiring a human worker in particular areas? But Chiang imagines that neon lights will enhance the human experience, not replace it; 15 years ago, no one used smartphones, and now these are constant portals that help us connect and improve our efficiency. "People have always feared that technology will replace (manpower in some) areas." But technology has always advanced: I'm a scientist, so I still think it's building technology as progress for humans," Chiang said. "If you have a tool to progress, why wouldn't you use it?" The second concern is whether neon lights can be used to impersonate people, a la Deepfakes, which is a completely different technology, STAR Labs says in an FAQ, and neon lights cannot be used to manipulate existing media. People cannot copy themselves or other people from Neons either; While some of his current creations are based on real people, the neon lights will ultimately be completely original. "We don't make replicas of people. It's not us," Chiang said. Which leads to the next question: will neon lights gain sensitivity, rise up, and replace humanity? "I'm a neuroscientist: we don't even understand the brain! How can you model something that you think will be even smarter than the human brain when we don't even fully know how it works? human brain?
Neon lights: make machines more humane to prevent humans from entering the machines
While introducing Neon to a host of media and technology professionals at the CES booth at STAR Labs, Mistry explained how her young daughter interacted with the smart assistant Alexa — she yelled at her. Because? Because intelligent assistants, like many AIs today, are shallow and submissive. They serve, help and do not impose. And the language we use to interact with them are pure commands. Why would we bother with the same consideration we use when talking to humans? It has an effect on us, says Mistry. "We are more like machines than machines more like humans," Mistry said. "Simply turning on speech in machines or in a human user interface will not make them human." With Neon, we want to break this barrier. We also want our conversations with machines to be more human. "These are STAR Labs' ambitions, and we'll wait to see if their execution corresponds to such lofty goals. They don't have a business case or implementation planned yet, and aside from the aforementioned beta, 2020 looks like a year of research and development." to see what people and industries expect from neon lights, and ultimately what neon lights can be capable of.Perhaps STAR Labs will share more when they launch their own event, Neon World 2020, though they haven't publicly announced exactly what it will be or when it will happen. Like the rest of Neon, it is largely a mystery box, which has earned it much scorn and criticism from the online press. But for his part, Chiang is not surprised. "It's funny, I read an article that many years ago people made fun of electric cars until Tesla came out and proved that it was useful. Sometimes you just have to be the one to introduce it," Chiang said.- Discover all the CES 2020 covers from LaComparacion. We're live in Las Vegas to bring you the latest technology news and releases, as well as hands-on reviews of everything from 8K TVs and folding screens to new phones, laptops, and smart home devices.