I think there will be a perverse benefit to Vancouver's historic lack of large-scale office buildings and general market orientation towards small- to mid-size tenancies because it will make it easier for companies to right-size their office space without running up against the leasing and subletting constraints of large floorplates. For building owners, while it's obviously ideal to have a large tenant take up some or all of a building, having the ability to accommodate subletting or floor-shedding without the awkwardness of trying to demise a very large floorplate building should help keep up leasable area efficiency.
Moreover, if we do encounter a wave of Class B and C office conversions to residential, and conceivably even some partial or total conversions of certain Class A product, the smaller floorplates are more readily adapted to the generally shallow configuration of residential uses and the need for daylighting of bedrooms. Though, on that note, there is precedent from the Yaletown warehouse residential conversions to permit interior bedrooms without daylighting, provided that the primary bedroom has this (either directly, or in an open concept loft-configuration). Perhaps there may be some accommodation for this in office to residential conversions? There would already been need to waive the requirement for suites to have outdoor living space (patios/decks).
The fairly diverse office tenancy of Vancouver should also help it weather sector-specific downturns that may result from broader economic inflection points, such as the anticipated reduction in employment for tasks that can be wholly automated or fundamentally reshaped through the adoption of AI. This is nothing entirely new, but like the elimination of 'typing pool' and 'secretarial pool', and near elimination of executive secretaries, the space needs and composition of the office were slowly but inexorably reshaped through the wholesale adoption of the computer, word processor, and means for direct communication (e.g. direct dial phones and multiple lines, personal printers, voicemail, fax, email, cell phones, smartphones with email, video conferencing, laptops, the proliferation of courier services, etc.).
With respect to so-called AI, someone described the current large language model (LLM) AIs as "word calculators" and that really resonated and stuck with me. I fully expect that portions of my job, for example, will soon shift to telling an AI what to write, editing the output, and using that in the course of creating my overeall work products. In much the same way I could spend hours doing long-hand arithmetic or plugging figures into a calculator, using Excel to manage information and perform math is just wildly more efficient. Likewise, I could, hypothetically, develop the skillset to manually manipulate images, plans, and maps, but Illustrator is wildly more efficient at this, too. I am still using the program to accomplish the task I want, but I'm not necessarily doing all the work on a number by number or pixel by pixel basis. A few hours spent in Illustrator creating a map could yield the same thing a talented draftsperson would have produced with pen and paper fifty years ago. But while that would have been their whole skillset, for me that is just one small part of the overall set of skills that I need to use to do my job. Of course it's frustrating that the productivity benefits that have accrued over the decades have not flowed, by and large, to the person that posses the required skills, but that's another story.
For my work, I think that AI-based tools will very soon be used to generate repetitive or high-level (i.e. not very important) blocks of text for a document. It will be a new tool that lets me save me time by offloading certain types of work onto a machine. If integrated with a credible search tool, as opposed to the current crop of AI's propensity to make shit up, I could also see it being contextually aware enough to populate and cite current figures and statistics as part of the generation of text. I think that the fundamental skill of operating a LLM AI tool will be being able to craft effective prompts. I would not be surprised if a very light-touch programming language were to emerge for 'power users' who want to get the most out of the tool, as opposed to the general public who may use it as a toy or engage at only a superficial level. Ultimately this will result in the creation of LLM AI APIs and simplified user interfaces that would let the programming language back end be interacted with in a more user friendly manner. Ultimately, I will still be operating the LLM AI as a tool that will let me do my job and keep up with the expectations of the skillset a professional in my field must possess.
__________________
VANCOUVER | Beautiful, Multicultural | Canada's Pacific Metropolis
Last edited by SFUVancouver; Jun 28, 2023 at 3:46 PM.
|
Prev
Linear Mode
