3 Reasons Why Slowing Down AI Adoption Is The Need Of The Hour
When “artificial” intelligence is at par with “human” intelligence, it’s time to hit the brakes!
Artificial Intelligence (AI) is being rapidly adopted across the globe, disrupting how we live, play, communicate and work — but you knew that already.
Traditionally, disruption has always had a negative connotation — an event that entails a radical realignment of thoughts, actions and beliefs.
Isn’t that what AI is currently asking of us?
A year ago, I would have lost my patience searching for a free image of a human being arm-wrestling an AI robot. Today, Lexica created one for me in the same time it took me to write the prompt.
We’re living in a time where a virtual bot can generate images, text, audio and video with human-like qualities in the blink of an eye.
This is our reality now.
In fact, this blog could have been written by yours truly, or a Large Language Model —there is no way of knowing.
Nothing wrong with that, you may say.
Well, at the moment, it may not be the case; but what happens when I start writing about the health benefits of smoking?
Smokers have a lower risk of developing some types of cancer, including lung cancer. — ChatGPT
What happens when AI chatbots become the preferred mode for research over search engines?
Fortunately, we’re not at that stage yet. But the faster we adopt AI, the faster businesses will churn out publicly-available AI tools that can spread misinformation, or be leveraged for malicious intents.
I recently wrote about the topic here:
In this blog, I will condense the issue into three simple reasons, explaining why slowing down the adoption of AI is the need of the hour.
Reason 1: The Fallacy That Technological Advancement Cannot Be Slowed Down
This misconception stems from an age-old marketing tactic known as first-mover advantage. Essentially, it says that a business is better off than its industry competitors if it’s the first to market a new product or service.
That’s the case with emerging technologies such as AI.
The business perspective is that if they don’t create and market it, someone else will. And if that’s the case, they might as well go ahead and do it.
Naturally, this is also a thinly veiled excuse to dominate a nascent market niche and rake in the big bucks.
Yet, the first-mover advantage doesn’t always work out.
If you’re over the age of 30, I’m sure the word “PDA” will bring back some memories. (For the Gen-Z readers, I’m not talking about that PDA!)
PDAs, or personal digital assistants, were palm-sized devices that can be thought of as the ancestors of modern-day smartphones. These nifty devices were all the rage in the late 90s!
Palm, Inc. had the distinct first-mover advantage with the Palm Pilot. Modern smartphone businesses such as Google, Apple and Samsung were nowhere to be seen — so why isn’t Palm a household name today?
As I said, the first-mover advantage does not guarantee success.
Palm failed to iterate and improve their product, and slipped entirely when Apple debuted the iPhone in 2007. As we know, Apple paced its innovations, slowly iterated and enhanced their flagship product to make it a mainstay on the market.
When it comes to AI, businesses need to slow down and iterate their offerings to address consumers’ pain points, rather than trying to be the first-mover in the industry.
The fact that AI innovation cannot be slowed down is a myth that AI companies perpetuate to keep pushing out half-baked products.
Unlike Palm, Inc., rushing to launch AI tools without safeguards will hurt consumers as much as the business — if not more.
2: We Are Setting Ourselves Up For An AI Arms Race
A picture paints a thousand words…
The phrase “AI arms race” has been gaining traction since 2010, and the two key protagonists here are the United States and China.
When asked whether the United States should slow down AI advancement after the debut of the revamped Bing on 10th February, Microsoft officials stated they couldn’t afford to as “we’re in a two-horse race between the US and China.”
The same sentiment was echoed by Senior Vice President and the Global Cyber Security & Privacy Officer (GSPO) for Huawei, creator of the world’s largest Chinese LLM.
This comes from two leading businesses invested in the AI sector. The threat of an arms race should never be the incentive to accelerate any technological innovation, especially Artificial Intelligence.
Funnily enough, the US and China both called for ‘responsible’ use of AI in military applications earlier this year.
Yet, this is not enough to entirely dismiss the notion of an arms race.
While actual use-cases may be decades away, no nation should be allowed to test and develop intelligent AI that can be integrated with their military services. Experts remain hopeful that the international security community will intervene and ensure a transparent exchange of dialogue between the two nations. However, that doesn’t always work out.
Look at this issue in context of the destruction of Hiroshima and Nagasaki. It was only after the world grasped the truly destructive abilities of nuclear weapons that we managed to achieve global nuclear non-proliferation.
With AI, we cannot take that risk — we need to slow down its unchecked growth before we find out its catastrophic potential with autonomous military applications.
3: The Illusion That We Cannot Control AI Without Advancing It First
This argument can be applied to every technology — in retrospect.
When camera-enabled smartphones hit the market in the 2000s, it was the start of a technological disruption.
Smartphone cameras enabled new innovations; video calling, QR code marketing, hi-res photography, capturing a video of your cat vocalizing, etc., using a device the size of your palm. It was revolutionary!
However, in the following decades we realized that this convenience came at the cost of our privacy. Today, smartphone cameras can potentially capture images, record soundbites or videos without permission to reveal sensitive and private information.
The argument that “we can only fix the issues with AI once we advance it” is a risky proposition, given that it will have much more far-reaching implications than smartphone cameras. After all, smartphone cameras can be covered up!
Learning from past mistakes, we have the opportunity to fix the issues with AI as we progress, albeit by moving a bit slower.
OpenAI’s ChatGPT garnered over a million users in 5 days — but several users have reported that it consistently produces incorrect, fake and misleading answers.
In the future, by the time we lock the stable, the super-intelligent AI horse will be freely galloping away.
What Tech Companies Need To Understand
We already have one excellent reason to slow down the adoption of AI — the fact that they are black boxes.
AI systems take in data, process it, and generate outcomes, but the “how” remains a mystery, even to the businesses that develop it.
It’s like watching a magic trick — the audience knows the reveal, but the exact method is reduced to guesswork.
When (not if) we develop a super-intelligent AI, we will undoubtedly connect it to the Internet, allowing access to every piece of information ever recorded. Without understanding how the AI understands and processes this information, blindly relying on its outputs would be reckless and potentially dangerous.
The black box issue, coupled with the three reasons listed above, clearly show that decelerating AI adoption in the short run is actually in the best interest of consumers and businesses alike.
