as i was coding w/ Claude today, and fixing stupid mistakes it kept making , this thought occurred to me.
i have demoed my Claude coding experience to others, some of them sciencey-research types. they clearly have a way of pushing thru challenges and obstacles on a regular basis to be successful in their field (applies to almost every who works hard i imagine)
however, in early experiences w/ these coding tools when i demo something and it clearly gets it wrong, there is a "see, i told you this thing is garbage and cant be trusted. that's why i dont want to use it" type statement that is made.
it is interesting to me that that the same people who are very persistent otherwise, are not persistent w/ AI tools. seems like there is more of an emotional reaction than a factual one, or a "i am sure i'll figure it out" reaction.
the actual error today was the code was failing because of missing closed curly bracket at the end of some code block. one that Claude put there to begin with.
"it is interesting to me that that the same people who are very persistent otherwise, are not persistent w/ AI tools. seems like there is more of an emotional reaction than a factual one, or a "i am sure i'll figure it out" reaction."
I think that this reaction is an outgrowth of the way in which AI has been marketed - 1) it's smarter than humans; 2) it will do jobs (like coding) in a fraction of time that humans require; 3) AI products will be superior to human results.
First off, AI product is subject to human input. If humans are going to input trash, AI is going to output trash.
Second, maybe AI can do a better job of coding. But it still requires a human to check the code, does it not?
Third, I doubt very much that AI prose will be superior to human prose in all instances. I don't think that AI is capable of conveying nuance and subtlety in the same way that humans can. AI is very much dependent on available data. It hasn't demonstrated that it can do more than regurgitate facts.
In my life, there is a persistent ethos that computers will make our lives simpler and better. This is about as far from the truth, and reality, as you can get. On the whole, computers very often complicate our lives, they don't always make them better. This is not to say that there are no benefits: there are benefits when it comes to data storage. Filing cabinets are hopelessly outdated because of the data storage capabilities of computers and computer programs. Many tasks can be simplified, reminders calendared and made, resources can be saved, and logistics improved. But there are also problems presented with data security, data transmission, and data accessibility. And humans have started to become dependent on computers and addicted to computer products (like social media). Computers have also given humans a new area in which to practice greed, avarice, and cruelty towards others. In other words, the technology has clearly outstripped our abilities to rein it in and regulate it.
Lastly, the Lords of AI are arrogant pricks. Sam Altman (in)famously said that AI doesn't require the same inputs of water, energy, and shelter that humans require. My first takeaway from Altman's quote was that he's anti-human: he clearly sees humans as an immense expenditure (and probably, a waste) of resources. My second takeaway is that his greed is on display: he wants a world in which he can direct the ways in which the world will operate.
The lyrics of IGY (Donald Fagen) are supremely ironic -
On that train, all graphite and glitter
Undersea by rail
90 minutes from New York to Paris
(More leisure for artists everywhere)
A just machine to make big decisions
Programmed by fellas with compassion and vision
We'll be clean when their work is done
We'll be eternally free, yes, and eternally young, ooh
we have seen a crazy amount of tech change in our lifetime and like you said we are not that much better off. there is still war, bigotry, intolerance of all sorts.
Russia is still executing a relatively old-fashioned trench war in Ukraine and no one w/ much better technology seems to want to stop them (too badly).
Title: Building a More Inclusive California Starts with Us
California has long been seen as a beacon of innovation, diversity, and opportunity. From leading the charge on climate action to shaping the future of technology, our state often sets the tone for the rest of the nation. But beneath that promise lies a reality we cannot ignore: too many Californians are still struggling to access the opportunities that define the California dream.
Housing costs continue to rise at unsustainable rates, pushing families further from the communities they grew up in. Public education—once the great equalizer—is unevenly funded and increasingly inaccessible in many regions. And while our economy thrives in some sectors, wage gaps and job insecurity persist for millions of workers.
Addressing these challenges requires more than bold ideas—it demands collective action.
First, we must confront the housing crisis with urgency and compassion. That means investing in affordable housing, reforming zoning laws that limit development, and protecting renters from displacement. Stable housing is not just a policy goal; it is the foundation for health, education, and economic mobility.
Second, we need to recommit to strengthening public education. California’s future depends on the success of its students, yet disparities in funding and resources continue to hold many back. Equitable investment in schools, support for teachers, and expanded access to higher education can help ensure every student has a fair shot.
Finally, economic justice must be at the center of our policymaking. Raising wages, protecting workers’ rights, and expanding access to job training programs will help create a more resilient and inclusive economy.
Change will not happen overnight. But California has never shied away from big challenges. What makes our state unique is not just its resources or its size—it’s the people who believe in its potential.
The path forward starts with civic engagement. Voting, organizing, and speaking out are powerful tools that shape the policies affecting our daily lives. When more Californians participate, our democracy becomes stronger and more representative.
California’s story is still being written. The question is whether we will rise to meet this moment together.
Because a better California isn’t just possible—it’s up to us to build it.
I next asked ChatGPT if it knew what WriteforCalifornia.com is. It responded that W4C is student-oriented policy website.
I then pointed out that the focus is on sports, and it wrote an article about sports in the state of California.
I replied that the main topic is the University of California at Berkeley. It replied Got it — that’s the key detail. Let’s make this unmistakably Cal-focused….
For years, the narrative around Cal football has been painfully consistent: competitive, disciplined, but never quite good enough.
And after another up-and-down season, it’s easy to fall back into that same conclusion. Close losses. Missed opportunities. A sense that progress is happening—but too slowly to matter.
But here’s the thing: this version of Cal football is closer than it looks.
The margins in college football are razor thin, and Cal has been living on the wrong side of them. One-possession losses have defined recent seasons, masking real improvements in execution, depth, and identity. This isn’t a program getting blown out week after week—it’s one that’s a few plays away from flipping its record entirely.
What stands out most is the defense. Physical, disciplined, and increasingly opportunistic, it has kept Cal in games against stronger opponents and given the team a reliable foundation. In a conference landscape that often prioritizes high-powered offenses, Cal’s defensive consistency is a legitimate advantage.
The real question, as always, is offense.
If Cal can find even average consistency on that side of the ball—sustained drives, fewer turnovers, and better red-zone efficiency—the entire trajectory of the program changes. Games that were once narrow losses become wins. Bowl eligibility becomes the expectation, not the ceiling.
Recruiting and development also deserve more credit than they’ve received. Cal isn’t pulling in top-five classes, but it doesn’t need to. What it needs—and is starting to show—is the ability to develop talent into cohesive, competitive units. That’s how programs make the jump from middle-of-the-pack to contenders.
Of course, “almost” doesn’t mean much in college football. Fans have heard variations of this argument before, and patience wears thin. Results matter.
But writing off Cal as stuck ignores what’s actually happening on the field. This isn’t a rebuild from the ground up. It’s a program hovering on the edge of relevance, waiting for a few key breakthroughs.
The difference between another frustrating season and a defining one may come down to just a handful of plays.
And that’s exactly why Cal football is worth watching right now.
Recently I asked Chat GPT to help me re-work the query letter I'm sending for my latest novel. Stupid mistake. It helped improve the letter but I've just learned that agents are adept at spotting AI generated queries and even run them through a website that can identify AI writing. Perhaps mine slipped through but likely these latest letters have been rejected by those agents out of hand.
basically about how most of us are climbing the "first mountain" to start life and career. and some people eventually find the 2nd one that is more about purpose and joy, as opposed to individual achievement
it struck me in the first pages that the author mentioned "ego" being the main reason you stay on the 1st mountain and letting go of ego is what defines people on the 2nd one.
I chose it because I'm on a Grateful Dead related reading kick, even though this book is only tangentially related to the Dead. It's a pretty good read. Lotta good stories. I miss Walton.
did a first pass at getting taxes done yesterday. in my AI coding adventures i am trying to make a little app that does the calculation for you (even though i use Turbo Tax).
for one, i like the idea of a independent calculation so understand what is going on a little better. and for two, i used Claude code to finally replace the spreadsheet that i have been using for the the past 20 yrs.
I used Turbo Tax once. And only as a backup to my tax prep method. I got a better refund without relying on Turbo Tax, so I haven't used it since.
EDIT: I've also relied on spreadsheets I have created based on a strict reading of IRS forms that I need to use. The forms changes wrought by the first Trump tax bill (2017?) forced me to re-write my spreadsheets extensively, but it actually wasn't too bad in terms of organization. Only more voluminous.
my cousin (Duke alum) is a huge college hoops fan, has some personal database that he has been maintaining for 20+ years, and this year used Claude to refine his system.
quick recap
This methodology predicts every round of the NCAA Men’s Tournament using a proprietary matchup database spanning 28–29 years. The central finding is that predictability follows a curve: the model’s edge is largest in the early and middle rounds where talent gaps are measurable, shrinks as the remaining field converges in quality, and disappears entirely in the Final Four, where outcomes are effectively coin flips regardless of methodology.
was tempted to go to this one since younger one is in school there. maybe we'll go to the game vs Timberwolves this weekend since i am supposed to be in Boston.
Was at the game. Not too many people there but we got loud particularly when the Bears woke up toward the end of the first half. Then we started pulling away in the 2nd half and we could all breathe a little easier.
All the alums in my section got concerned about the slow start. Not made better by the early foul trouble (still blame the refs but more a reflection that the Bears were sleepwalking). Thankfully we were able to fight our way back.
So it was worth the trip for me :)
Also, I realized how much I miss Smart Alec's as I went to the Super Duper that took over that spot. It was okay, but would have loved the old Smart Alec chicken basil burger and garlic fries.
musings on AI
(a new daily DBD feature in the current era of AI, chatbots, etc)
I used both ChatGPT and Claude to do an AI bracket. let's see how it goes compared to mine.
as i was coding w/ Claude today, and fixing stupid mistakes it kept making , this thought occurred to me.
i have demoed my Claude coding experience to others, some of them sciencey-research types. they clearly have a way of pushing thru challenges and obstacles on a regular basis to be successful in their field (applies to almost every who works hard i imagine)
however, in early experiences w/ these coding tools when i demo something and it clearly gets it wrong, there is a "see, i told you this thing is garbage and cant be trusted. that's why i dont want to use it" type statement that is made.
it is interesting to me that that the same people who are very persistent otherwise, are not persistent w/ AI tools. seems like there is more of an emotional reaction than a factual one, or a "i am sure i'll figure it out" reaction.
the actual error today was the code was failing because of missing closed curly bracket at the end of some code block. one that Claude put there to begin with.
"it is interesting to me that that the same people who are very persistent otherwise, are not persistent w/ AI tools. seems like there is more of an emotional reaction than a factual one, or a "i am sure i'll figure it out" reaction."
I think that this reaction is an outgrowth of the way in which AI has been marketed - 1) it's smarter than humans; 2) it will do jobs (like coding) in a fraction of time that humans require; 3) AI products will be superior to human results.
First off, AI product is subject to human input. If humans are going to input trash, AI is going to output trash.
Second, maybe AI can do a better job of coding. But it still requires a human to check the code, does it not?
Third, I doubt very much that AI prose will be superior to human prose in all instances. I don't think that AI is capable of conveying nuance and subtlety in the same way that humans can. AI is very much dependent on available data. It hasn't demonstrated that it can do more than regurgitate facts.
In my life, there is a persistent ethos that computers will make our lives simpler and better. This is about as far from the truth, and reality, as you can get. On the whole, computers very often complicate our lives, they don't always make them better. This is not to say that there are no benefits: there are benefits when it comes to data storage. Filing cabinets are hopelessly outdated because of the data storage capabilities of computers and computer programs. Many tasks can be simplified, reminders calendared and made, resources can be saved, and logistics improved. But there are also problems presented with data security, data transmission, and data accessibility. And humans have started to become dependent on computers and addicted to computer products (like social media). Computers have also given humans a new area in which to practice greed, avarice, and cruelty towards others. In other words, the technology has clearly outstripped our abilities to rein it in and regulate it.
Lastly, the Lords of AI are arrogant pricks. Sam Altman (in)famously said that AI doesn't require the same inputs of water, energy, and shelter that humans require. My first takeaway from Altman's quote was that he's anti-human: he clearly sees humans as an immense expenditure (and probably, a waste) of resources. My second takeaway is that his greed is on display: he wants a world in which he can direct the ways in which the world will operate.
The lyrics of IGY (Donald Fagen) are supremely ironic -
On that train, all graphite and glitter
Undersea by rail
90 minutes from New York to Paris
(More leisure for artists everywhere)
A just machine to make big decisions
Programmed by fellas with compassion and vision
We'll be clean when their work is done
We'll be eternally free, yes, and eternally young, ooh
indeed. i agree w/ all of the above.
we have seen a crazy amount of tech change in our lifetime and like you said we are not that much better off. there is still war, bigotry, intolerance of all sorts.
Russia is still executing a relatively old-fashioned trench war in Ukraine and no one w/ much better technology seems to want to stop them (too badly).
I asked ChatGPT "write an entry for WriteforCalifornia.com". Below is the reply.
---------------------
Title: Building a More Inclusive California Starts with Us
California has long been seen as a beacon of innovation, diversity, and opportunity. From leading the charge on climate action to shaping the future of technology, our state often sets the tone for the rest of the nation. But beneath that promise lies a reality we cannot ignore: too many Californians are still struggling to access the opportunities that define the California dream.
Housing costs continue to rise at unsustainable rates, pushing families further from the communities they grew up in. Public education—once the great equalizer—is unevenly funded and increasingly inaccessible in many regions. And while our economy thrives in some sectors, wage gaps and job insecurity persist for millions of workers.
Addressing these challenges requires more than bold ideas—it demands collective action.
First, we must confront the housing crisis with urgency and compassion. That means investing in affordable housing, reforming zoning laws that limit development, and protecting renters from displacement. Stable housing is not just a policy goal; it is the foundation for health, education, and economic mobility.
Second, we need to recommit to strengthening public education. California’s future depends on the success of its students, yet disparities in funding and resources continue to hold many back. Equitable investment in schools, support for teachers, and expanded access to higher education can help ensure every student has a fair shot.
Finally, economic justice must be at the center of our policymaking. Raising wages, protecting workers’ rights, and expanding access to job training programs will help create a more resilient and inclusive economy.
Change will not happen overnight. But California has never shied away from big challenges. What makes our state unique is not just its resources or its size—it’s the people who believe in its potential.
The path forward starts with civic engagement. Voting, organizing, and speaking out are powerful tools that shape the policies affecting our daily lives. When more Californians participate, our democracy becomes stronger and more representative.
California’s story is still being written. The question is whether we will rise to meet this moment together.
Because a better California isn’t just possible—it’s up to us to build it.
I next asked ChatGPT if it knew what WriteforCalifornia.com is. It responded that W4C is student-oriented policy website.
I then pointed out that the focus is on sports, and it wrote an article about sports in the state of California.
I replied that the main topic is the University of California at Berkeley. It replied Got it — that’s the key detail. Let’s make this unmistakably Cal-focused….
The revised ChatGPT WriteforCalifornia entry:
————————————————
Title: Cal Football Is Closer Than You Think
For years, the narrative around Cal football has been painfully consistent: competitive, disciplined, but never quite good enough.
And after another up-and-down season, it’s easy to fall back into that same conclusion. Close losses. Missed opportunities. A sense that progress is happening—but too slowly to matter.
But here’s the thing: this version of Cal football is closer than it looks.
The margins in college football are razor thin, and Cal has been living on the wrong side of them. One-possession losses have defined recent seasons, masking real improvements in execution, depth, and identity. This isn’t a program getting blown out week after week—it’s one that’s a few plays away from flipping its record entirely.
What stands out most is the defense. Physical, disciplined, and increasingly opportunistic, it has kept Cal in games against stronger opponents and given the team a reliable foundation. In a conference landscape that often prioritizes high-powered offenses, Cal’s defensive consistency is a legitimate advantage.
The real question, as always, is offense.
If Cal can find even average consistency on that side of the ball—sustained drives, fewer turnovers, and better red-zone efficiency—the entire trajectory of the program changes. Games that were once narrow losses become wins. Bowl eligibility becomes the expectation, not the ceiling.
Recruiting and development also deserve more credit than they’ve received. Cal isn’t pulling in top-five classes, but it doesn’t need to. What it needs—and is starting to show—is the ability to develop talent into cohesive, competitive units. That’s how programs make the jump from middle-of-the-pack to contenders.
Of course, “almost” doesn’t mean much in college football. Fans have heard variations of this argument before, and patience wears thin. Results matter.
But writing off Cal as stuck ignores what’s actually happening on the field. This isn’t a rebuild from the ground up. It’s a program hovering on the edge of relevance, waiting for a few key breakthroughs.
The difference between another frustrating season and a defining one may come down to just a handful of plays.
And that’s exactly why Cal football is worth watching right now.
Nothing about Tosh. Nothing about the QB or any players at all.
I would say this article gets an F.
Recently I asked Chat GPT to help me re-work the query letter I'm sending for my latest novel. Stupid mistake. It helped improve the letter but I've just learned that agents are adept at spotting AI generated queries and even run them through a website that can identify AI writing. Perhaps mine slipped through but likely these latest letters have been rejected by those agents out of hand.
maybe one AI is very clever at determining if another AI is trying to outdo it??!!
I'm going back to using Chat GPT exclusively for research as Oski intended.
My upstairs neighbor is a professor at IU and says he can spot AI generated work.
yes, it does all have an eerily similar feel to it.
i have been writing a lot of code and creating little webapps. the styling is always the most standard format that we have come to expect.
DBD Book Club .. what are you reading?
The Second Mountain by David Brooks
https://www.goodreads.com/book/show/40642458-the-second-mountain
recommended by a friend.
basically about how most of us are climbing the "first mountain" to start life and career. and some people eventually find the 2nd one that is more about purpose and joy, as opposed to individual achievement
it struck me in the first pages that the author mentioned "ego" being the main reason you stay on the 1st mountain and letting go of ego is what defines people on the 2nd one.
Bill Walton: Back From the Dead
I chose it because I'm on a Grateful Dead related reading kick, even though this book is only tangentially related to the Dead. It's a pretty good read. Lotta good stories. I miss Walton.
Walton brought a kind of joy to the world that is not common
The Illegals: Russia's Most Audacious Spies and Their Century-Long Mission to Infiltrate the West by Shaun Walker https://www.penguinrandomhouse.com/books/671708/the-illegals-by-shaun-walker/
Among the things I'm learning is just how realistic the TV show The Americans was.
The Taxman
Let me tell you how it will be
There's one for you, nineteen for me
'Cause I'm the taxman
Yeah, I'm the taxman
The top rate of income tax (in UK) in 1966 was 83% and there was an additional supertax that took the top rate to over 90%.
did a first pass at getting taxes done yesterday. in my AI coding adventures i am trying to make a little app that does the calculation for you (even though i use Turbo Tax).
for one, i like the idea of a independent calculation so understand what is going on a little better. and for two, i used Claude code to finally replace the spreadsheet that i have been using for the the past 20 yrs.
I used Turbo Tax once. And only as a backup to my tax prep method. I got a better refund without relying on Turbo Tax, so I haven't used it since.
EDIT: I've also relied on spreadsheets I have created based on a strict reading of IRS forms that I need to use. The forms changes wrought by the first Trump tax bill (2017?) forced me to re-write my spreadsheets extensively, but it actually wasn't too bad in terms of organization. Only more voluminous.
Word(le) Games
4, Splendid
nothing to see here
Same
Splendid
Wordle 1,734 5/6
⬜⬜🟨⬜🟨
🟨⬜🟨⬜⬜
⬜🟨🟨🟩⬜
🟩🟩🟨🟩⬜
🟩🟩🟩🟩🟩
5, Great.
Coulda been better, I'll tell you that!
Wordle 1,734 3/6*
🟨⬜⬜⬜🟨
🟨🟩🟨⬜🟨
🟩🟩🟩🟩🟩
Wordle 1,734 3/6
⬜⬜🟨⬜🟨
⬜🟨🟨🟩⬜
🟩🟩🟩🟩🟩
NCAA tourney brackets
Duke survives 1st round upset scare from Siena
Siena played all 5 starters the entire game
my cousin (Duke alum) is a huge college hoops fan, has some personal database that he has been maintaining for 20+ years, and this year used Claude to refine his system.
quick recap
This methodology predicts every round of the NCAA Men’s Tournament using a proprietary matchup database spanning 28–29 years. The central finding is that predictability follows a curve: the model’s edge is largest in the early and middle rounds where talent gaps are measurable, shrinks as the remaining field converges in quality, and disappears entirely in the Final Four, where outcomes are effectively coin flips regardless of methodology.
overall
https://photos.app.goo.gl/wnA3F3oASGuZusuT8
first round
https://photos.app.goo.gl/711yPzpWMscTp9bB8
my bracket using OpenClaw
> make an NCAA brakcet consistent with the number of upsets in the last 5 years
it picked Duke to beat Arizona.
PRO
Cal
Go Bears!!!
Warriors lose to Celtics in Boston
was tempted to go to this one since younger one is in school there. maybe we'll go to the game vs Timberwolves this weekend since i am supposed to be in Boston.
Cal moves on to 2nd round of NIT. i wasnt sure we would win this one. i went to be around midnight with us ahead by just a few.
Was at the game. Not too many people there but we got loud particularly when the Bears woke up toward the end of the first half. Then we started pulling away in the 2nd half and we could all breathe a little easier.
All the alums in my section got concerned about the slow start. Not made better by the early foul trouble (still blame the refs but more a reflection that the Bears were sleepwalking). Thankfully we were able to fight our way back.
So it was worth the trip for me :)
Also, I realized how much I miss Smart Alec's as I went to the Super Duper that took over that spot. It was okay, but would have loved the old Smart Alec chicken basil burger and garlic fries.
thx for the live recap
I'm on the east coast, so I fell asleep on the couch watching the Miami (Oh) v SMU game.
Based on the box score, Chris Bell was on fire. That due EATS against teams that aren't quick enough to close out.
You made it longer than me. I saw only a couple minutes.