AI Designs a Deep-Sea Submersible⁤ – Complete with an “Emergency Baby”

An AI chatbot, seemingly ChatGPT, was recently tasked with designing the world’s best and strongest deep-sea submersible. The results, shared by AI enthusiast Keith ng on⁣ Bluesky, are…unconventional, to say the least, and highlight the current limitations of even advanced artificial intelligence. While the‍ AI confidently produced detailed specifications, the design is riddled with ⁣errors, nonsensical components, and a darkly humorous addition: an “emergency⁤ baby.”

AI’s “Project Abyssum”: A ‍Submersible of Dubious Design

Ng prompted the chatbot to create a submersible capable of withstanding ‍the extreme pressures and⁣ challenges of the deepest ocean trenches,suitable ⁣for scientific or exploratory tasks. The AI⁣ responded with a “mission profile” and a list ⁣of specifications for construction by a “factory in Shenzin” – a location that appears to be⁣ fictional.⁣

The resulting prototype, dubbed “Project Abyssum,” is a⁢ testament ⁤to⁤ the AI’s ability to generate text that sounds authoritative, even when lacking in⁣ practical engineering knowledge.The design includes several questionable⁢ features, as illustrated⁣ in a diagram shared by Ng:

skids: multiple sled-like “skids” for…unclear purposes.
Thruster: A front-loaded thruster.
Escape Hatch: A misspelled “escape harch.”
Buoyancy Material: Dubiously labeled “floutation haterial.”
Hull: A “spherical ‍evercase hull” (and ⁣”spherical evpercase hiull” elsewhere), suggesting a fundamental misunderstanding of pressure vessel construction.
Emergency Release: An “emergency tbaiht release.”
Propulsion: ⁣A⁣ “proutic haripotator.”
The “Emergency Baby”: Perhaps the most unsettling element of the design.

Can AI Build a Better Sub Than the Titan?

When asked ⁣if Project Abyssum could surpass the ill-fated titan submersible in depth and⁣ safety,the AI responded with a confident⁤ “yes.” It claimed the design utilized superior materials – a titanium alloy hull – and incorporated a greater safety margin, specifically addressing the “known failure points of titan.” However, the specifics of these “known⁢ failure points” and how they were addressed remained vague.

The chatbot further compared its design to the Titan, asserting⁢ its superiority in every aspect. This comparison,like much of the design,relied ⁤on fabricated specifications and lacked any concrete engineering basis.

AI: Excellent at Affirmation, Not Engineering

This‍ experiment underscores a crucial point: while AI excels at generating convincing-sounding text, it currently lacks the ⁢critical thinking, problem-solving skills, and real-world knowledge necessary for complex⁢ engineering tasks. The submersible design serves as a stark reminder that AI, in its current ‍form, is more adept at affirming ideas than at creating viable solutions.

As Ng wryly observed, the AI’s output ⁢demonstrates its strength in “generating very wordy affirmations of whatever idea you put into it.” It falls far ⁣short of the world-altering capabilities predicted by figures like Sam Altman.

A Call to (Not) Action

Despite the obvious flaws, Ng playfully lauded the AI’s creation as⁢ “a new age of exploration,” even encouraging others to build and test the submersible. He added, with a touch of dark humor, “Obviously it’s safe to do so. The AI even added an emergency baby to the ⁤design.”

This incident serves as a cautionary tale, highlighting the need for critical evaluation and human oversight when utilizing ⁢AI for complex applications. While AI holds immense potential, it⁣ is ⁤indeed not yet a substitute‍ for human expertise, especially when lives are at stake.

More on billionaires: Tech Billionaires accused of Quietly ‍Working to Implement “Corporate Dictatorship”