James Higginbotham

11.3K posts

James Higginbotham banner
James Higginbotham

James Higginbotham

@launchany

API and AI coach. Author of "Principles of Web API Design" (Addison Wesley). Instructor. Coach. Consultant. Enterprise architecture

Colorado, USA Katılım Mayıs 2009
890 Takip Edilen3.1K Takipçiler
James Higginbotham retweetledi
Vaughn Vernon
Vaughn Vernon@VaughnVernon·
We software developers habitually use CRUD and collection terminology in our designs/code. It's a knee-jerk decision because our programming facilities—frameworks and libraries—are chock-full of them and we want to go fast. Our brains become wired to that "language" and we start to think that every abstraction is just a variable, container, or disk/database. Pro Tip: Challenge yourself and your team to reject CRUD and collection terminology and seek out more fitting words directly from the business.
Vaughn Vernon tweet media
English
0
10
27
1.9K
James Higginbotham
James Higginbotham@launchany·
@MIT_CSAIL “A bad vibe coder is someone who never looks either way before crossing a highway” - me
English
0
0
7
301
MIT CSAIL
MIT CSAIL@MIT_CSAIL·
"A good programmer is someone who always looks both ways before crossing a one-way street." — Doug Linder
English
10
87
790
35.3K
James Higginbotham
James Higginbotham@launchany·
Developers went from buying a laptop or desktop and some dev tools like a pro IDE and commercial libraries, and coding as much as they want with it… …to being told that we should instead use AI to write our code, turning AI into a utility with peak and off peak pricing that we will have to manage like a Nest thermostat, and coding agents that allow vendors to monetize your personal downtime while it runs Ralph loops. And developers helped build those tools as if they were asked to train their replacement without knowing it
English
0
1
1
273
Steve Ruiz
Steve Ruiz@steveruizok·
It does bother me that it costs money to code now
English
193
168
3K
177.7K
James Higginbotham
James Higginbotham@launchany·
@ChiefEngineerCE I’m seeing this now with a very large enterprise. The replacements don’t understand security and privacy implications of their work and behavior. It won’t be pretty in another year or so
English
0
0
6
236
Chief_Engineer
Chief_Engineer@ChiefEngineerCE·
The 40 Percent Rule: How Organizations Quietly Lose Competence There is a rule of thumb that has been passed around operations and turnaround circles for decades. It is rarely taught in business school anymore, but those of us who have watched organizations get gutted by private equity or aggressive cost, outsourcing, and offshoring have seen it play out repeatedly. We are all witnessing it ...because Financial folks aren't aware of it and think a warm body is all an org chart needs. It starts with layoffs justified for financial reasons. About three months later the first cracks appear. The experienced people who kept systems running through undocumented workarounds and tribal knowledge are gone. The new staff assumed those tasks were obvious. The organization hobbles along at first, but the clock has started. Then the 40 percent rule kicks in. Any organization that replaces more than 40 percent of its core technical and operational personnel within a 3 yr period, typically crosses into intellectual bankruptcy roughly two years later. By year three the loss of institutional know-how becomes irreversible. The company breaks down into silos that spend most of their time firefighting until competitors take their customers or the organization slowly decomposes. This pattern is not theoretical. It has been observed across multiple large corporations over the last century whenever finance or HR fully takes over hiring and firing decisions. Amazon is currently tracking this trajectory in real time. Microsoft appears to have passed the point several years ago. From an engineering perspective this is straightforward systems failure analysis. Tacit knowledge - the undocumented glue that holds complex operations together, cannot be transferred or recreated at the same speed it is removed. When you lose too much of it too quickly, the remaining staff no longer have enough context to rebuild what is missing. The decay then becomes self-reinforcing. This explains why so many organizations that looked healthy on paper suddenly begin missing deadlines, burning cash, and losing market position two to three years after major workforce reductions. Downtime events, meetings to say- we cant do that any longer (as if it were intentional). The balance sheet improved for now. The institutional competence did not. Yes the balance sheet always improves because it takes a while for customers to find a new source and for the new companies to form to take the business. Engineers usually see the warning signs first because we are the ones forced to keep undocumented systems alive long after the experienced people are gone. Real question for the engineers, operators, techs and turnaround people reading this: What was the earliest operational signal you noticed when your organization crossed the 40 percent threshold? Comment if you have experienced similar below. No names needed - just the signs you saw on the ground. Bookmark this if you have watched a once-strong company slowly lose its ability to execute after heavy cuts. Tech Wednesday
English
71
171
935
100.9K
James Higginbotham
James Higginbotham@launchany·
Sure thing. And thanks for working toward on-peak and off-peak character generation as we move toward AI-as-a-utility. Those of us that wrote software character-by-character are very excited to treat compute like electricity 🙄
Sam Altman@sama

I have so much gratitude to people who wrote extremely complex software character-by-character. It already feels difficult to remember how much effort it really took. Thank you for getting us to this point.

English
0
0
0
135
James Higginbotham
James Higginbotham@launchany·
@straczynski Same. I’m heading to the finale once again, with TKO up next. I’d say this is my 10th or 11th rewatch since the first airings in syndication and that finale gets me every time
English
0
0
0
23
James Higginbotham
James Higginbotham@launchany·
Fascinating to see “team and enterprise AI twitter” dispel myths about MCP being dead to “indie AI twitter”
English
1
0
2
107
James Higginbotham
James Higginbotham@launchany·
@BallerToy1327 Loved Murray as a recurring gag. For me, it is tied with BtVS when Spike was a “cousin of Xander-ers” in season 4
English
0
0
2
28
James Higginbotham
James Higginbotham@launchany·
@JiltedValkyrie I just started a rewatch, given all the discussion around Babylon 5 lately. Currently 8 episodes into season 1
English
0
0
2
24
Jilted Valkyrie
Jilted Valkyrie@JiltedValkyrie·
First ever watch of Babylon 5 starts tonight.
Jilted Valkyrie tweet media
English
27
2
96
2K
James Higginbotham
James Higginbotham@launchany·
@yongfook Then watch as agents game your pricing by upgrading and downgrading multiple times per minute if you use tiered pricing. Value based pricing plus payment registration with optional spending caps might be essential moving forward
English
0
0
0
14
Jon Yongfook
Jon Yongfook@yongfook·
Am I crazy to think that upgrade / downgrade / payment should all be exposed by API in your app, so AI agents can buy and self-upgrade? Is this already a thing?
English
48
6
135
27.3K
Frank Turek
Frank Turek@DrFrankTurek·
What book of the Bible are you reading right now?
English
4.2K
168
2.8K
206.4K