
The picture was created with the assistance of Microsoft Design
Are you able to think about that within the close to future, massive language fashions will grow to be a serious innovation device?
Certain, why not?
I envision LLMs being more and more engaged in producing new concepts and fixing issues by analyzing massive quantities of information and figuring out patterns that people won’t be capable to see.
If we imagine — and I do — that innovation is about connecting the proverbial dots, then what may be higher at seeing extra connectable dots than an LLM?
However for now, we’re caught with doing innovation in an old school means: by interacting with different human beings.
That’s the place the position of connecting human “dots” (a.okay.a. networking) within the innovation course of comes into the highlight.
In one of my earlier posts, I wrote a few profound unfavourable impact on innovation attributable to Prohibition, a nationwide ban on the sale of alcoholic drinks in 1920–1933.
Earlier than the passage of federal prohibition, U.S. states and counties may decide for themselves whether or not or to not enable alcohol consumption in bars and saloons. When federal prohibition went into impact, counties that had been beforehand moist noticed an 8–18% drop in patenting relative to constantly dry counties in the identical state.
In an excellent 2020 paper, “Bar Speak: Casual Social Interactions, Alcohol Prohibition, and Invention,” Michael Andrews exhibits that the impact of Prohibition on innovation had nothing to do with alcohol consumption per se.
As an alternative, Prohibition affected innovation by disrupting pure social networks.
Previous to Prohibition, bars and saloons acted as social hubs through which people may alternate data in an off-the-cuff setting. When individuals stopped going to bars, that they had misplaced a “platform” for exchanging concepts. The injury to the innovation course of materialized virtually instantly.
It’s fascinating to see parallels between Prohibition, an ill-conceived authorities regulation, and the COVID-19 pandemic, a pure catastrophe. (I’m not a subscriber to a concept that COVID-19 has been a Frankenbaby born in China’s Wuhan Institute of Virology.)
And but, each did precisely the identical: they disrupted pure human networks, Prohibition by stopping individuals from socializing after work, and COVID-19 by abrupt shifting to distant work.
A latest paper took a take a look at what occurred to worker collaboration when Microsoft shifted, firm-wide, to distant work in the course of the first six months of 2020.
The paper exhibits that the shift to distant work brought on enterprise teams inside Microsoft to grow to be much less interconnected, rendering the collaborative networks extra static and siloed.
Working from dwelling, Microsoft workers did proceed to collaborate with others across the agency — utilizing all out there digital communication instruments — however the sample of collaboration modified. They had been spending extra time working with their sturdy ties however much less time interacting with their weak ties.
That is vital as a result of sturdy ties are higher suited to transferring data, whereas weak ties are extra possible to offer entry to new data.
In different phrases, distant work led to a sample of collaboration that’s much less prone to lead to producing novel, probably modern concepts.
It have to be confused that the timeframe of the examine didn’t enable evaluating the impact of regret work on innovation per se — solely on the sample of worker collaboration.
Nevertheless it’d be naive to anticipate that the disruption of established collaboration networks attributable to the abrupt shift to distant work is not going to trigger a long-lasting unfavourable (“Prohibition-style”) impact on innovation at Microsoft — and at different tech corporations as effectively, for that matter.
Not too long ago, we’ve been seeing growing makes an attempt by corporations to get away from distant work and produce workers again to workplace — all beneath the banner of restoring the broken sample of worker collaboration.
On the floor of issues, this is sensible. The “return-to-office” coverage may be considered as an try, whether or not aware or not, at rebuilding broken company innovation networks and bringing them, as shut as doable, to the pre-pandemic state.
Sadly, this sample of rebuilding — “forceful” is the phrase that involves thoughts first — carries a threat. By and enormous, company workers are sad with the necessary return to places of work. In line with a survey carried out by Make clear Capital, 68% of workers mentioned they’d quite search for a brand new job than return to the workplace.
This is perhaps extra a menace than actual motion. However let’s not neglect {that a} highly effective driving drive of innovation is freedom. Sure, you possibly can drive individuals again to their places of work and make them talk in particular person. However will these sad individuals freely alternate artistic concepts with different sad individuals?
I’ve my doubts.
In addition to, though I totally recognize the position serendipitous encounters might play within the innovation course of — some extent many executives invoke when pushing for return to workplace — I’m not satisfied that company innovation should a lot rely on an opportunity bumping of individuals into one another.
Can we innovate at scale whereas ready for a spark of magic taking place at a watercooler or in a toilet?
After all, I’m all for individuals assembly their colleagues frequently (which doesn’t essentially imply every single day, does it). However do we actually want places of work for that?
Why gained’t we convey workers to off-site conferences as an alternative?
Inserting workers in pleasant settings — with ample time for bodily exercise and rest — will generate situations for a free and unrushed circulation of concepts between the individuals. This can be a higher likelihood of them producing modern concepts than speaking to one another over a cubicle wall.
In addition to, we have to complement in-person interactions with a platform that may assist worker collaboration no matter workers’ bodily presence within the workplace.
Such a platform ought to meet two primary necessities:
- Be always-on. Company innovation struggles when run as a collection of “occasions” following one another, usually with an extended pause between them. As an alternative, company innovation have to be a steady circulation of parallel innovation initiatives with fastened begins and finishes, in order that a minimum of a number of are all the time lively at any time.
- Be project-based. With regards to innovation, individuals ought to have causes to collaborate, not simply “alternate concepts.” That is greatest achieved by making use of the agency’s innovation potential to fixing particular, strategically vital to the agency, issues.
There isn’t any want to begin from scratch: a prototype for such a platform already exists.
I name it inside innovation networks, and I’ll return to this matter in my subsequent piece.



