Waymo, a Google-owned company, analysed the safety performance of its autonomous vehicles recently. Photo/123RF.
A couple of months ago I was asked; “Could I produce this column using AI?”.
“I don’t know why I would, but that’s an interesting thought,” I replied.
Then a couple of days ago I was confronted by a colleague who said; “Don’t ever produce your column using AI”.
So that did it!
So long as my by-line is at the top of this column, it will continue to be produced by my writing it out long hand, then being dictated into a computer program- which struggles with some words, and then being edited before posting Tuesday, for Thursday’s Northern Advocate edition. That I promise.
But this AI thing creeps up on you.
Last week I was at a board meeting where we approved a policy - “The Use of AI Within the Workplace”.
With permitted uses being: Data analysis and reporting; Drafting documents emails and meeting notes; Summarising internal communications and policies; Assisting with scheduling and task management; and Supporting decision making through structured insights.
There’s a lot more detail than that of course, but it really opened my eyes as to where the use of artificial intelligence (AI) might be heading.
Then there was an article in last Sunday’s Sunday Star Times headed up “Love or theft? What do Kiwi creatives think about generative AI?”.
Included were some comments such as; “He isn’t against the use of AI, he just prefers wrestling with ideas and concepts and finding solutions where you learn who you are, and what your opinion is” and “If you give over the creative process to AI, then you are not the writer anymore. You are the editor, you are the person doing all the bits and pieces, but you aren’t the author”.
AI, it seems, is well regarded as a tool to help and stimulate the creative process, but the final product is owned by the author.
A bit like this column really, various comments being cherry picked from wherever, then coalesced into some directed thinking.
But it’s in the area of driving and self-drive cars that AI has greater relevance to the thrust of these columns, and it was fascinating last week to watch local business owner, Lindsay Faithfull, drive a self-driving, newly equipped Tesla.
The development of self-drive cars has had its own supporters and sceptics for decades now, but hovering over it is the recognition, that humans behind the wheel have fallibilities, and maybe robots could drive us better.
It is hard for some of us to accept that we are almost certainly worse drivers, than we think we are. We drive distracted, we drive drowsy, we sometimes drive angry, and worst of all, we drive impaired more often than we should. As well, our brains take a while to engage when faced with an unusual situation. The answer for some of this it seems, is with AI.
In a recently released, peer-reviewed study published by the journal Traffic Injury Prevention Waymo, a Google-owned company, analysed the safety performance of its autonomous vehicles over the course of 56.7 million miles driven in Austin, Los Angeles, Phoenix and San Francisco.
All without a human safety driver present, to take the wheel in an emergency. They then compared that data to human driving safety over the same number of miles on the same kind of roads.
Compared to the human driver the Waymo, self-driving cars had: 81% less airbag deployed crashes; 85% fewer crashes with suspected serious or worse injuries; 96% fewer injury crashes at intersections; and, 92% fewer crashes that involved injuries to passengers to pedestrians. Now this is a substantial company study on American roads, and New Zealand roads and conditions may bear no comparison.
It is significant then, that Tesla in the last month, has released a right-hand drive version for New Zealand and Australian roads, as was publicly driven by Lindsay Faithfull last week.
Tesla labelled this technology as, “full self-driving supervised” to be used only by a fully attended driver.
The system is designed to make driving easier, less stressful and safer. It has 360-degree visibility, checking blind spots, negotiating intersections and roundabouts, traffic lights and stop signs, and reacting to pedestrians, cyclists and other vehicles.
The system though, also monitors driver attentiveness with cabin cameras which cannot be turned off.
If the cabin camera does not have clear visibility of the driver’s hands and arms locations, a message is sent to the driver. If the driver continues to not pay attention or ignores prompts, the supervised self-drive is turned off. If the driver does not resume manual steering, warning flashes are turned on and the car comes to a complete stop.
Everyday driverless cars seem a long way off for the average New Zealand driver. But the AI phenomenon is with us now and closer to us than we think. We need to ensure and treasure the fact that we can continue to think. That’s the beauty of the human brain.