If I were a betting man, I’d put >20% odds that this or a successor technology reaches nearly ubiquitous deployment in the first world within 10 years.
It doesn't matter what a PM at Google tells you; they'll rotate outwards after next perf cycle and it will be a new PM who will be utterly powerless to stop Google's mismanagement from shutting down the project you rely on.
You have to treat all non-core products as ephemeral.
This would almost be a subtweet except that it's Google and by now this is just tired market commentary, but the instigating event for it is Google shuttering Stadia, their internal teams learning with ~hours of notice, and external partners learning from Twitter.
There were a lot of software developers who bet fairly big on making engineering investments so that the promised horde of Google users would buy their games and, if you are in a position to make this decision in the future, get hard commits from Google up-front in contracts.
I’ve estimated that, across all customers, this feature will save many engineer *lifetimes* of work which was previously necessary to ship e.g. SaaS products but which produced no unique value.
“Customers want to be able to change their subscription or update their credit card”, “CS says people lose emailed invoices all the time and want to download them all from one place”, “It turns out users in Germany have extremely specific requests for language on invoices”, etc.
Nobody went into engineering because they really wanted to sweat the details on an invoice listing screen.
We liberate people from that so that they can do things for their customers which uniquely benefit from their skills.
One of the many places where widespread deployment of great general purpose hardware (smartphones) is enabling software to deliver experiences which previously required specialized hardware (card readers).
This probably reads as most interesting for casual commerce, because card readers historically imply a degree of organizational infrastructure that the smallest firms don’t have, but I suspect over the long term it actually gets used more for intrinsically mobile transactions.
For example, any sort of service provider who physically visits a customer should be able to capture payment on the same app that they use to record the fact of work being done, without needing a dedicated tool to do so.
I have so blacked out the packaged software portion of my entrepreneurial career that I was about to write a followup tweet saying “Or so I’ve been told” and then remembered “Oh yeah I suffered the alternative *for years*.”
Underappreciated outside of the industry and extremely appreciated by the guy whose personal motto is Charge More: the transition to SaaS accompanied a large increase in software prices (spread over time) which more closely rationalized the price to value ratio.
Playing with Stable Diffusion (via diffusionbee.com ) and while I think "best thing I've ever done on a computer" is a high bar it definitely, definitely hits that for the kids.
Lillian: "Can I have a fairy that looks like me?"
Me: *tries to guess how the Internet would phrase this* "Portrait of Japanese pixie with gossamer wings, cherry blossoms in background, DeviantArt"
Jumping off this anecdote to mention: many bureaucracies will phrase “We lost it” as “You failed to submit it” until you show them Certified Mail, Return Receipt Requested.
The IRS has lost, found, then lost again my e.g. tax returns before, and every time they complained my accountant pulled out the receipt and said Not My Problem.
From this follows two observations:
a) Since you know how the game is played, any important submission to a bureaucracy goes over Certified Mail Return Receipt Requested.
b) Any mail tagged with that reads to every mailroom as A Dangerous Professional Wrote This.