Opinions expressed by Entrepreneur contributors are their own.
Last year many of us spent time thinking about the problem of AI bias carefully described by one of the authors of “Coded Bias”, the famous one Netflix documentation. Now that another boost in generative AI’s popularity is set to continue, job replacement talks are back in the game.
It’s one of the most detailed reports on how AI could potentially automate (or, many fear, replace humans in their skilled jobs). Goldman Sachswhich was vehemently spread under a multitude of alarmist headlines 300 million potentially replaced jobs around the world.
Some of the reported data in particular suggests that 18% of work worldwide likely to be computerized and the impact on more developed economies could be worse than that in emerging markets, for example.
Curiously, the recent generative AI boom has coincided with several back-to-back waves of layoffs in the online tech industry, making just a sort of minor panic in countless discussions around the internet even more understandable.
See also: The 3 Principles of Building Anti-Bias AI
However, the report itself suggests that the so-called “suspension to automation” in no way implies the elimination or removal of the labor involved by humans. More importantly, many of the non-employee occupations are not even prone to negative effects.
On a larger scale, according to some experts, the ability to run next-gen AI technology will be critical for the pros, rather than being rendered obsolete by chat-GPT-like solutions any time soon. As Ingrid Verschuren, head of data strategy at Dow Jones called“Humans are the actual “machine” that powers AI.”
Face the reality behind the hype
So, as Goldman Sachs estimates, until almost 25% of all work could be managed by AI completely in the next few years. But what exactly does that mean for a specialist in the legal department, a copywriter or a motion designer, for example? To tell the truth, not that much.
A friend of mine who runs a video production studio has been testing AI solutions for generating images for a while now, and it turns out it’s been quite an arduous journey scraping the creative inspiration out of the machine learning algorithms the whole time. The default images are often a bit generic (and often somber), so their design team couldn’t actually apply the newly acquired AI-assisted assistance to any significant extent.
Meanwhile, in newsrooms, the recent trend of running ChatGPT queries in relation to some news personalities and seeing the not-so-truthful results also proved that the point of truthfulness is the weakest point of Generative AI.
And given so much false narrative and how easily the generative AI tools can be persuaded (e.g. writing content with non-existent facts when given in the assigned request), I highly doubt their legal advice is qualified enough to do so They with, let alone replace even an inexperienced but hungry paralegal for their software equivalent.
Will the future sustain our fears?
While the current state of Generative AI is obviously not as advanced as its founders would like us to believe, some of the job forecasts for 2024 may seem overly pessimistic. Of course, there’s a chance that technology will have a significant impact on our workforce in one way or another over the next decade. So how can we be prepared?
Here are some key areas entrepreneurs might want to keep an eye on:
Don’t rush into cut-offs
Whatever niche you are in, the current state of Generative AI does not have the skills and competencies to replace any of the skilled specialists on your team.
More importantly, even as AI advances further, you’ll likely still need your team to manage the new software (i.e. explain exactly what to do and then verify the result) to get the best results .
Some of the most vivid examples are code reviews/optimizations, editing the scripts created by AI, re-reviews of accounting and engineering projects, and medical physical exams/prescription reviews, but the list is virtually endless.
See also: History has shown what happens to companies that shy away from new technologies. So why are so many afraid of Generative AI?
Check your facts
While we leave it to the media and celebrities to worry about the possible negative effects of complex deepfakes made possible by the introduction of generative AI upgrades, using ChatGPT or similar tools to search for information remains a very tricky proposition.
As the training of the algorithms evolves, the risk of getting completely misguided will definitely decrease, but chances are we won’t be able to trust the AI generated text/image any time soon.
Even if this aspect will remain of primary importance in editorial offices, law firms and political offices, all calculations provided by the advanced machine learning algorithms must also be checked again, at least in the selected data cohorts.
Notably, the amount of time and operational resources that are inevitably required to conduct these reviews/audits actually challenges a common belief that the expanded use of AI leads to greater productivity at lower budget expenditures.
Beware of the bias
The first thing we learned from ChatGPT’s launch was that the latest “knowledge acquisition dates to 2020 – 2021”, but the most important thing is that despite its latest upgrades, the generative AI is still dated or shall we say biased.
Here are some examples to prove my point.
I ran a simple query asking ChatGPT to “tell me a story about two people,” and what I got was a cheesy rom-com about John and Mary. Then I ran a quick query in the appropriate generative AI software to draw me two people on the beach and I got an image of two men (although the scene structure was no doubt good). Presumably, after analyzing my query, the algorithm “decided” that “people” should primarily mean “male people”.
What this means for entrepreneurs using Generative AI, whether they work in a creative industry or not, is that they must not only have a clear understanding of AI bias risks, but also have a willingness to triple check and then to update the results generated by the intermediate software before they are integrated into one of the further work products.
Prospects for 2023-2024
Long story short, whatever the misconceptions we may have about generative AI at this point, they probably won’t remain relevant 10 years from now. However, the most sensible approach to its use remains in moderation. In simple terms, overdoing its benefits will definitely be harmful, but over-focusing on its possible effects can be just as much.
To quote Ms. Verschuren from Dow Jones, it’s still up to us humans to determine our future and tweak our machines for better results, no matter how complex.