Google’s Kozyrkov and Alexa’s moxie both impress me – what’s your take?
Should you believe any technology today is artificially intelligent? There’s a battle shaping up over standardized interfaces for voice interface technologies. Who’s going to lead the defacto standard?
Two stimulating developments:
- Cassie Kozyrkov, the head of Decision Intelligence at Google, wrote another brilliant blog post: Artificial Intelligence: Do stupid things faster with more energy! According to her, the scary part of AI isn’t the robots. It’s the people. I’ll explore some of her thinking and advice below.
- Amazon, along with thirty other vendors – including Harmon, Microsoft, Salesforce, and Sonos – announced the Voice Interoperability Initiative. They set a simple initial goal. Devices should support many “wake words” – such as Alexa or Cortana – letting users talk to the virtual agent service of their choice in a secure manner and switch from one to another at will. What’s right and wrong with this?
Is AI intelligent?
Let’s start with Cassie’s blog post. The heart of her argument is that there is no such thing as an “artificial intelligence.” The idea of building an “artificial intelligence” is a ruse, an artifice, a conceit or a provocative assumption. She doesn’t quite say it that way but she and I are of a like mind here.
The scary part of AI is not the robots. It’s the people.
AI is not a monster. People are. AI is not intelligent. The people who write the code, design the models, tag the data, run the training sessions and test the outcomes — they provide the intelligence! The machine is a dumb brute, born of human-generated or selected algorithms and data.
I recently wrote about this issue on Quora:
Which do you think is more likely: that AI will save us, or that it will ultimately destroy us? Why?
We are AI. We created it. It’s not really very smart but what it can do under limited conditions is amazing. Will we continue to create ever more amazing technologies? Yes, that’s what AI is, amazing new technologies. Will we continue to learn their weaknesses (yes, even though people don’t like to focus on the shortcomings — the strengths, hyped strengths — are more seductive and enticing.)
We build AI. We marvel at it. We break it down. We build newer AI to replace previous generations of AI. The birth-growth-maturation-decline-death and replacement cycle continues.
The question you’re asking is will we save ourselves or kill ourselves? Will we continue to build autonomous (automated) weapons, for example?
The human race comes closer to killing itself every day. We are changing our planet, our only planet, and not all for the good! We continue to wipe out millions of people in senseless wars and genocides. We commit mayhem. We burden people with religious myths that drive them to kill and maim (and enrich the few behind the frenzies.)
We are losing a sense of truth versus fiction. We dis science. We disdain vaccination for our children. And on and on.
Can we destroy ourselves with technology? Yes. Will we? What do you think? This is up to people!
People shouldn’t fear AI. They should fear evil or incompetent people! And they should stop thinking about animating spirits (“AIs”) unless they’re writing scary books or screenplays for scary movies.
Another quote from Cassie:
If you ask me whether I’m scared of AI, what I hear you asking me is whether I am scared of human negligence. That’s the only way the question makes sense to me. My answer? No. I’m optimistic about our AI future. I believe that people can learn to build safe and effective systems responsibly, driving progress and making life better for those around them.
Voice Interface Standards.
I just said to my iPhone “Alexa, put English Muffins on my shopping list” and my phone responded (in text, not voice) “Notes hasn’t added support for that with Siri.”
I recognize that the Voice Interoperability Initiative (VII) is aspirational. They have the right opening aspiration: let smart devices recognize many different wake-words and invoke whichever underlying service the user, at the moment, wants to be using.
But VII is not getting off the ground unless they can get either Apple or Google to play nicely and support the VII in a reasonable and timely fashion.
I’ve been writing about the age of conversational systems for more than three years and tracking its evolution for the last seven. For the industry to move into the age of conversational systems, we’re going to need a de facto standard interface to these systems!
So kudos to Amazon here — but, as yet, they have no real (voice invocation) presence on smartphones or tablets, even though they are trying.
For me (and the industry as a whole, including customers) this comes down to a horserace between Amazon, Apple, and Google. Yes, there are at least 30 more firms I could mention — including a number in China — but let’s keep the list focused for now.
(Let’s not forget about Chana. I consider it unlikely that the Chinese will support an American de facto standard here and equally unlikely that American firms will support a Chinese de facto standard. Wonders may occur and prove me wrong, but it’s not likely in the next few years even though I hope for it.)
De facto wins
I predict the solution will come from the de facto leader in usage and integration with other underlying services and applications in the consumer market. (There will be no de jure or negotiated standard here.)
Who will win the battle for standardizing the interface to voice interfaces? How will that struggle play out? Do either Google or Apple cave early? Which one? Why? And when will it happen?
Punters, state your positions and place your wagers! Let the intellectual gamesmanship and betting commence.
Disclosure: The views and opinions in this analysis are my own and do not represent positions or opinions of The Analyst Syndicate. Read more on the Disclosure Policy.