Sagie Davidovich on the future of AI and ethics


Stuart Kennedy
Contributor

Artificial intelligence is moving beyond just predicting the future to shaping it to our will, although ethics are extremely important to this brave new world.

That’s according to Sagie Davidovich, the chief executive and co-founder of Israel-based AI leader, SparkBeyond. Mr Davidovich will be speaking at the virtual Australia-Israel Innovation Summit on October 21.

SparkBeyond came about in 2013 when the AI company was founded in Israel by Mr Davidovich and Ron Karidi. The company’s IP was forged when the founders did a moonshot experiment to trawl the internet for open source code and build an AI machine from all the available open source algorithms.

The platform that resulted can generate millions of hypotheses per minute and grind through hundreds of good and bad ideas every second.

Abstract Tunnel
Bending the future: Artificial intelligence ventures need ethical frameworks

The machine has taken on all sorts of problems from retail intelligence to controlling cancer.

In a cancer fighting use case, the SparkBeyond system was able to help doctors increase their ability for early detection of colon cancer. The AI machine explored millions of ideas and discovered intricate patterns related to changes in haemoglobin levels over time which pointed at colon cancer potentially developing.

Mr Davidovich said the SparkBeyond works by rapidly generating and testing hypotheses. “By combining pieces of code, and then testing the hypotheses that are created on the day and using the data to prove or disprove each hypothesis.

“For example, there was a large retail chain in Asia, where they want to drop in 5000 new store locations, the machine found that proximity to laundromats correlates with the profitability of the retail store. That’s one of our favourite examples, because you can create the heat map of where to locate the stores.”

There’s predicting the future, and then there’s shaping the future, which is where SparkBeyond diverges from other analytics engines, according to Mr Davidovich.

“This is the divergence point between SparkBeyond and the rest of the analytics world. Everyone’s focus is prediction. We say, maybe we can change the system, so that whatever others are trying to predict, we can actually shape,” he said.

“When you when can invent a lightning rod, why do you need to predict where the next lightning strike is going to hit? It took us several years to articulate and crystallize this. But essentially, what we realized is that we help overcome cognitive bottlenecks and cognitive biases that we as humans have in ideation and problem solving.”

In bad news for gambling outfits, SparkBeyond operates on a strict ethical profile as much as it does on the revolutionary AI engine.

“We developed a moral compass that was created collaboratively by the entire company,” said Mr Davidovich. “It acknowledges that the world is messy. It’s not about yields. It’s about principles.”

The company colour grades types of problems it is interested in assisting with, from green to red with blue, orange and grey zones in between. The red zone lists types of organisations SparkBeyond chooses “not to work with into the foreseeable future”.

These include companies that provide or promote gambling, the tobacco industry, the weapons industry, pornographers and political parties.

In the green zone, SparkBeyond lists organisations with which it will proactively seek to licence its platform. These include organisations whose primary focus is sustainable development such as sustainable agriculture outfits, government agencies and NGOs dedicated to education, planning, environmental protection, reducing poverty and unemployment, smart city initiatives, curing disease, fighting human trafficking and fire and rescue services.

In the blue zone sit insurance, retail and financial services. Orange zone companies are those where SparkBeyond will licence its platform to but “with a view to influencing them to progress socially beneficial initiatives.”

These include oil and gas energy companies, mining operations, alcoholic drink producers and games companies. The grey ethical areas include government agencies such as defence, intelligence, secret service and immigration.

“It’s important for companies to understand that AI is much more than technology, AI creates a very different culture around experimentation and turns the entire company to be data driven. The second thing that’s part of this has to do with ethics,” said Mr Davidovich.

Do you know more? Contact James Riley via Email or Signal.

Leave a Comment

Your email address will not be published.

Related stories