Elephant in the context window

Ask not what your elephant can do for you, ask what you can do for your elephant.

There is a famous Indian tale about a group of blind sages who bump directly into an elephant. None of them can see the elephant as a whole, and so each of them tries to extrapolate a part within immediate tactile reach to the rest of the body. Nor can a single blind sage be entrusted with thoroughly palpating the entire elephant and successfully describing it: he or she may have never seen an elephant before, and may have never even heard of elephants. Moreover, this person may not have a context window large enough to persistently remember all the surfaces encountered during the process in order to build a truly coherent view.

Also, what if there is more than one elephant to be described, or perhaps an elephant and a rhino? Such a situation may require the sage to consciously manage the context window, zooming in at a particular spot or out to grasp a larger area, while patiently recording the accumulated data to figure out what is really happening.

Hollywood occasionally helps us to understand such phenomena. If you observe the “Wall Street” protagonists via an expanded context window, you will recognize them to be near-identical to characters in “Pretty Woman,” “The Wolf of Wall Street,” and “American Psycho“. A few particular scenes dangerously overflow into “The Irishman,” while others connect optimistically to “Back to School.” Together, these movies construct for us a rather consistent view of the American business, technology and political landscape of the 1980s and ’90s.

Technology-wise, engineering and manufacturing IT was born in the 1970s with the advent of MRP systems, followed by CAD of the 1980s, and then by the PLM systems of the 1990s, evolving over the next 30 years into a sort of a digital elephant herd.

  • Elephants became a natural phenomena of our daily lives, whether it’s a camouflaged pocket-size elephant from Ukraine, or an American aerospace OEM‘s massive herd of elephants with different locations and rules.
  • The modern engineering view of an elephant is overwhelmingly 3D, while the ERP people believe the elephant is flat. The C-Suites firmly consider the elephant to be a financial construct that improves the bottom line, and they couldn’t care less about its shape. Meanwhile, the corporate IT team braces for yet another root canal procedure on the elephant under minimal sedation.
  • The bottom line improvement requirement is a function of the elephant(s) being continuously healthy with predictably affordable total cost of ownership, and perpetually growing productivity across the room.
  • Predictable affordability is easier said than done. Companies which consciously monitor their own context window, and process an avalanche of information coming from vendors (software architecture and deployment footprint, releases roadmap and general attitudes) and expert veterinarians can make better decisions about their particular herd composition. They can always hedge their bets, for example with open data standards such as STEP or JT.
  • Productivity is as dependent on the invisible technological sophistication as, ultimately, on bringing that technology to the end-users’ fingertips in the most ubiquitous, simple and reliable manner.

Enterprise search is a particularly important engineering and manufacturing productivity track. Specifically, everything hinges on the ability to quickly and conveniently zoom in on and out of any particular piece of any 3D elephant in terms of levels of abstraction and relationship density.

With the right approach and competence, the sages miraculously gain eyesight and not only can see the whole elephant, but are also able to turn it around and upside down, move it in the direction they need, and exploit it for their wise purposes.

We at Senticore have always been interested in that topic of elephant research. Over the last year, we have worked quietly with our partners on technology we call “AI-Enabled 3D Context Discovery”. We think of it as a pretty unique MRI-like approach built on top of graph technology and generative AI for complex and discreet product design, manufacturing and MRO processes dealing with 3D models and various associated content.

I would compare using our solution to deal with engineering data to watching the movie “Wall Street” with eyes and brains being used at 100% capacity: the movie quickly exposes traditionally obfuscated forces right into the regular folks’ TV screens with extreme aesthetic and intellectual clarity. Our solution allows us to fetch engineering 3D-centered data into an explicitly-defined context window at breathtaking speed.

Just like “Wall Street” exposes glimpses of tsunami emerging from behind the boardroom curtains, our concept uncovers the dangers of electrical harnesses hiding behind metal sheets. It becomes possible to understand how to drill through these surfaces safely, and how to inspect them reliably. Just like “Wall Street” and its related movies hint at a certain path to society’s salvation, our concept allows designing SAE and IEC standards-compliant products easier.

Speaking of salvation, fixing Western civilization circa 2024 might be similar to Thornton Melon passing the exams and performing Triple Lindy in “Back to School”: exceptionally difficult and extremely satisfying. I can say the same about watching our concept of AI-enabled 3D Context Discovery developing into a great tool for the engineering and manufacturing domain.

How are your elephants today? Give us a call to schedule an examination.