By Christine Persaud

If the future LG presents about artificial intelligence (AI) in the home becomes reality, it could be both tremendously convenient and terrifying at the same time. The company has identified four levels for qualitative jumps in AI capabilities, which Dr. IP Park, President and CTO, outlined during the company’s press conference this morning, alongside Jean-Francois Gagne, co-founder and CEO of Element AI.

The first level of this AI experience (AIX) is efficiency and is characterized by the simple ways AI allows us to better automate device operations and user interaction. The best example is voice control and products like smart speakers that help facilitate our interactions with smart devices in the home. It could also be devices like smart air conditioners with smart sensors that can detect when people are in a room and adjust the temperature accordingly. This is pretty garden variety AI and commonplace nowadays thanks to platforms like Amazon Alexa, Google Home, and Apple HomeKit.

The second is personalization or pattern learning involving how AI can accumulate data through interactions with these such devices and build a memory. An example might be a robotic vacuum that can learn where it got stuck before because of obstructions or a tight corner, for example, and avoid those areas the next time you run an automated cleaning session.

Jean-Francois Gagne, co-founder and CEO, Element AI

The third is where things get creepier: causality learning and reasoning. This is when AI not only knows what to do, like wake you up in the morning at a certain time, but also understands why in order to create better outcomes for you. It’s reasoning by intelligence, and involves devices collaborating with one another to share what they learn and relate insights for causes and effects. “AI is more about the collective intelligence of a system,” said Gagne. Research is just starting around this, but it could involve a wake-up alarm not only ringing at a set time, but then reminding you not to sleep in if you try to snooze it because you have a big meeting in the morning. Then it might recommend a good power breakfast to get you going, notice you’re nervous about an upcoming presentation and try to calm you by reminding you about a celebratory dinner that evening and even recommend good places to go.

Level four is where we delve into seriously innovative but also scary territory. Referred to as the exploration level, this is where AI begins its experimental learning phase. It creates hypotheses and proactively seeks new information. It’s like AI becoming a scientist, says Dr. Park, using scientific methods to become smarter and figure out what’s best for you. Optimization becomes a more personal process, supported and facilitated by AI.

Dr. IP Park, President and CTO, LG Electronics

The demonstrations reminded me of the USA Network series Mr. Robot, where one isolated character began to see Alexa as her “friend.” She would have conversations with it at home at night, ask the AI intense questions about life and love (to which it gave robotic and meaningless answers), and use it for everything from playing music to simply “someone” to talk to. Finally, a real-life friend smashed the unit against a wall telling her this device was not her friend.

Is AI’s intention to become so smart that we, as humans, can rely on AI to help us through tough times, recommend good restaurants, and get us through our days without any real human interaction? Of course, the hope is that this isn’t the case. In one demo, LG showed how the AI recommended to a woman who was nervous and stressed about her day that she should call her good friend for further reassurance. Still, do we really need AI to remind us to interact with real people? Or validate our feelings?

It’s one thing when AI can act as a personal assistant, keeping track of busy calendars, intelligently controlling our smart home devices so we can save time, energy, and money, and spend more time with families and friends. But when AI starts becoming a “friend” on which you rely on for more than that, it’s a frightening thought.

LG’s tagline this year is “Anywhere is home,” with its ThinQ experience now encompassing all AI involvement across its product lines. Home being anywhere is great, as is AI being with you wherever you go – as long as humans are, too.