August 2024
Reading time: 5 minutes
OPINIONS EXPRESSED ARE MY OWN
and do not reflect my employer
Dissimulation : to hide under a false appearance.
Synonyms: deceit, deception.
It truly is the future. AI drives your car. AI does your job. AI is sexting your girl. AI is playing Valorant on your computer. AI is wearing your shoes. Your morning bowel movements has a new, revolutionary, instant slide-out, no wipe AI technology from Google!
Ok, not really. In fact there are too many claims about what AI does.
Companies are misusing AI, by deceiving customers about the true nature and capabilities of their product or services.
It is dissimulation.
Companies will claim that their products are completely, 100% "AI-powered", creating an impression of cutting-edge, futuristic technology that is better than before, even if the AI used is rudimentary or not even AI but rather repackaged solutions.
Tech startups and established companies use AI as a selling point to attract media hype, investors, and customers. They insert "AI" into their branding, product names, and live demonstrations that seem incredible, too good to be true.
In reality, the "AI" that was promised can be a simple automation software that doesn’t involve any machine learning or neural networks. There are also companies that just make completely fake demos, to then later reveal that what they were demoing was what they “hope” their product will be able to do in the future. Google comes to mind with their Gemini AI model.
This is like demonstrating a tear-proof condom you are pitching to investors, only to later write on the condom box that you hope in the future the condom will be tear-proof.
The financial insentive
Companies are taking existing software solutions and re-marketing them as "AI" without making any significant changes or upgrades. Traditional data analytics tools have been re-marketed as “AI analytics” tools to tap into the everlasting hype of machine learning.
A lot, and I mean a lot of customer service chatbots are labeled as AI, but often they are mostly decision trees or scripted responses.
Startups are emphasising AI innovation and incorporation of it in their products to inflate their perceived value to investors.
Companies, especially in the tech industry, suddenly claim to be an "AI company". AI has magically upgraded their company to a new and better company. They do this without substantial change to their core business or even any solid evidence of how this was achieved with AI.
"It just works better now because of AI. Trust us bro."
John Startup - Calculator AI Incorporated
These pivots are strategic financial moves to align with investor interests, market hype, and expectations to innovate, rather than a genuine shift towards AI-driven advancements.
Fleetwood Mac - Little Lies
Companies are claiming that their systems use AI to enhance user experiences, improve decision-making, or that their entire company is AI-run, but they don’t provide any detail about how their AI works. No transparency makes it difficult to discern if AI is genuinely being used or if the claim is superficial and hollow.
A great example of this is The Amazon Go stores. They were from the very get-go marketed as using advanced AI and revolutionary machine learning to create a "Just Walk Out" shopping experience. The selling point was that customers could grab the products they wanted and just leave without going through any checkout. No card swipes, no cash, no Apple Pay, just walk out.
That’s crazy! Cool! AI is so amazing!
It was later revealed that instead of relying on AI to confirm purchases, Amazon, one of the richest companies in the world, had workers in India, one of the lowest-paid labor forces in the world, manually reviewing and correcting the system’s errors.
Oh...
These Indian workers were checking footage and data the whole time to ensure that Amazon's amazing AI had correctly identified products that the customers took off the shelves. They were supposed to be serving as a backup to the AI system. It seems however that humans were doing most of the work.
Some reports claim that up to 70% of all purchases had to be manually corrected with human intelligence and decision making.
So perhaps the AI was so advanced that it knew it wouldn’t have to do any real work because humans would correct the work it did anyways? I think not.
This human intervention directly contradicted Amazon's own claims that it was an autonomous AI-driven shopping experience. The company used AI as a buzzword to create the perception that their machine learning product and algorithms are more advanced than the actual reality; Indian low-wage workers doing what were essentially Amazon Mechanical Turk tasks.
The Carpenters - We've only just begun
Companies attributing the outcomes of their processes to AI without offering explanations can be quite dystopian and ruthless.
The fact that companies can simply hide behind a mask of AI with no transparency creates a great rift between company and customers.
In recent years, financial institutions like banks have said that they have deployed AI models to determine loan approvals when in actuality it might just be an excuse for declining loans to undesirable customer profiles that are less profitable for the institutions.
Whether or not they are using real AI is debatable, we will never know without transparency.
AI will obviously become better, but if people have to question what is and what isn't powered by real AI it will cause even more distrust towards corporations and core institutions.
Lack of transparency leaves the door wide open for companies to deceive and manipulate not just possible investors who will put their hard-earned money in their stocks but also the customers, who might not know that whatever service they are using is not actually AI, it’s just an Indian guy doing a task manually for breadcrumbs.
Dissimulation : to hide under a false appearance.
Synonyms: deceit, deception.
Show your support for more articles like this
