“Never trust anything that can think for itself if you can’t see where it keeps its brain!”
– Arthur Weasley, Harry Potter and the Chamber of Secrets
When OpenAI released ChatGPT to the world at the end of 2022, it seemed like magic come true. At last! Here was a machine that could think for itself and do the work of many (intelligent) humans in a matter of seconds. Tremendous power at little or no cost. The world could never be the same again; an invisible but real threshold had been crossed.
ChatGPT was soon followed by dozens of other LLMs (Large Language Models). The term “LLM” describes what the technology is and does:
– Large: These systems contain massive amounts of training data, consisting of billions of parameters and adjustable mathematical weights.
– Language: They are designed specifically to process, understand and generate human language.
– Model: They are mathematical frameworks trained on text to identify patterns and predict the next most likely word in a sentence.
That was a sober but rather boring description. Most of us prefer the much juicier term “AI”.
What’s in a name?
Unfortunately, referring to these new machines as “Artificial Intelligence” hides more than it reveals.
The “A” in “AI” is real.
The “I” in “AI” is fake.
Despite the lofty rhetoric from newly minted AI billionaires and their praise-singers, a machine can never acquire intelligence. However, it can achieve enough computational power to impress most of us, which is enough. Calling it “SI” (Sham Intelligence) would have been closer to the truth.
How Does This Affect Integrow Health?
As a 21st-century e-commerce business, we believe in using technology to help more people more efficiently (and, need it be said, more effectively). Machines run our website, machines run our software, and machines run our payment systems. There’s nothing new or unique about that. However, we also need an AI Policy to guide our use of this new kid on the tech block
When the South African government had to draft a policy on the use of AI, the responsible committee got AI to write it for them (probably in a few seconds). It is unlikely that any of the committee members bothered to read it, since nobody picked up on the fake references. Fortunately, intelligent people read the document intelligently and pointed out how fake the policy was.
Amusing as this little episode may seem, it points to the true danger of AI: We, the users, are dumbing down. We’re subletting our thought processes to dumb machines. How much critical thinking will be left in society even a year from today?
At Integrow Health, our basic approach to technology has always been simple:
Machines at the back, people at the front.
In other words, use machines to empower our staff, so they can help you better.
AI will not change this policy, but that’s easier said than done. We have a little AI chatbot on our website to help users when we’re not available. However, wherever possible, we will strive to communicate with you in person.
Machines break, software glitches and AI hallucinates. Maybe it is best to compare these 21st-century tools to a wild bronco recently tamed. Unless you remain vigilant and intentional, you’re likely to end flat on your face. As a beast of burden, the technology at our fingertips has almost unimaginable potential for good. As a god, this technology will destroy our humanity. Artificial intelligence, wrongly used, has the potential to make us Really, Really Stupid. Now wouldn’t THAT be a Faustian bargain of epic proportions?
You’ve probably figured out we don’t use AI to write our newsletters. The reason is simple, and it highlights one of the biggest drawbacks of AI (not that the fawning media will draw attention to it). That drawback is that all public LLMs have been “conditioned” to respond within predetermined parameters. When it comes to questions of health, the AI responses will inevitably point to an allopathic solution. AI does not bite the hand that feeds it. Some clever prompting may manage to dig a bit deeper, but you’re fighting an uphill battle. AIs like ChatGPT and Claude are not neutral when it comes to health and disease. Obviously, AIs should not be giving health advice. Guard rails are essential. However, when it comes to holistic health, AIs are of limited use.
In short: Enjoy our 100% organic, handcrafted emails! We won’t insult your intelligence by sending you computer-generated content.
In the future, we may use tech tools to help us serve you better, but we will never use tech tools to make any critical or clinical decisions. No matter how human-like AI interactions may become, we will not use that as an excuse to stop serving you (in person) with real gladness and diligence. At the end of any interaction, there will always be a human ready to brighten up your day.
Only Real Humans can love and be loved.
To your (intelligent) health!
The Team at Integrow Health
