Hunger Games for Robots
Prompt: humans are meat-based APIs authorized by the IT department (Midjourney)

Hunger Games for Robots

After customer discovery with HR and engineering managers, I was on team which pitched a business idea (modeled after the non-profit Graduate Management Attendance Council and GMAT exam).

The product and service was essentially a proctored computer adaptive test which provided a baseline to evaluate technologists on a wide set of skills.

With the benefit of hindsight, it had a whole set of issues, including self-selecting the talent pool, perpetuating biases, etc.

However, my memory of the experience was the reaction of an engineering professor tasked with chaperoning the SharkTank-like showcase.

He wasn't a judge, but ended up telling us how he really felt about the (hypothetical) business idea.

After noticing the judges were struggling to grok the problem being solved, he exclaimed that what the service was really solving for: cheating by students/early-stage employees!

That was the latent challenge being addressed on behalf of hiring managers.


Current Service

After summer vacation ends, ChatGPT use increases. Good!

As of last fall, a company is selling an “AI detection bypasser” to students.

More offerings like this will inevitably pop-up.

The product strategy will be to tease with a freemium model which starts with the same AI detection used by schools.

If the student’s text is flagged, they can upgrade to pay for AI to further paraphrase.

This iterates until the detector shows no machine influence.

If you're wondering what is the human, societal, or pedagogical problem being addressed, you're probably not alone.


Future Service

Seeing around corners, I imagine there will soon be multiple VC-backed companies selling a feature to schools which detects if a student has used an “AI detection bypasser” service.

It may not even be necessary, with current technology, the end result is probably rendered into gibberish text.

Employing expensive computing resources to train language models which learn human sentences and then have the machines declare Mortal Kombat on each other smells like an example of infinite regress .

Turtles…all the way down.

The one-upsmanship to avoid learning and skills development seems like self-sabotage.


Gamesmanship

Education curriculums and problem-solving skills are being applied to each other in a form of machine-augmented intellectual sport.

Forget Zuck vs Elon, the next billion-dollar e-sport might involve paying customers clamoring to watch virtual influencers fight for survival in a digital cage match.

Unfortunately, there are pre-existing examples of pattern-matching investors funding technical talent to build products devoid of sense-making or value.

You'd want to hope that Hunger Games for Robots? doesn't become a reality, but remember the Yo app was a thing .

要查看或添加评论,请登录

社区洞察

其他会员也浏览了