The punch line, yes of course I voted for Trump. Just in case you want to stop reading now :). I will have more to say about this below, because I voted FOR Trump, not against Harris, although I think she's a horrible candidate and would have voted for virtually anyone else. I am the classic 'used to be a "centerist" Democrat but now find myself a "far-right" Republican'. So in case you care, here's how I voted: ==== NATIONAL President / Vice President: Donald Trump / JD Vance. I like both of them, more on this below. US Senator: Steve Garvey. This is more of an anti- Adam Shiff vote, but I do like Steve Garvey, and not just because he was an excellent first baseman. He's a good positive speaker and his positions make sense.
==== STATE (California) Measure 2 - $10B school bond: No. I don't like specifically authorized expenditures, even when I agree with the target. BTW, California already has $78B in debt. Maybe if we stopped paying for illegal immigrants and DEI programs we could afford to support public education. Measure 3 Right to marry regardless of sex: Yes. This is a silly waste of time and money, since same-sex marriage is already legal due to federal court rulings. Measure 4 Bonds for water, wildfires, climate protection: No. As noted, I'm not a fan of earmarked spending, and on top of that I'm not a fan of "climate protection" spending. So double no. Measure 5 lowers local bond threshold: No. This is a meta-spending measure, it reduces the approval requirement for local bonds from 2/3 to 55%. Why? So that we can have more local bonds for "affordable housing" and "infrastructure". Those are laudable targets for government support but should be paid for out of budgets, not bonds. So no. Measure 6 removes involuntary servitude: No. Currently being in jail or prison means you can be forced to work. So be it. Measure 32 raise minimum wage: No. I feel minimum wage hurts the people it is designed to help. We need more lower paid jobs for more people. Not to mention, it's inflationary. Oh, and California has to compete with the rest of the world, which might not have a minimum wage. Measure 33 expands rent control: No. Keeping government out of markets is the best policy. Like minimum wage, this hurts the people it is designed to help, by discouraging new housing construction. Measure 34 additional regulation for federal drug discount program: No. The federal drug discount program is already market meddling; this measure seemingly adds a layer of additional meddling to restrict spending on housing. Measure 35 make tax on MCOs permanent after 2026: No. Wow this one was complicated to understand. Apparently there is a tax on MCOs which took effect in 2009 and has been renewed several times, and it currently expired in 2026. Seems like it's been in place for 17 years, so be it, it can be renewed as it has been, doesn't have to be made permanent. Measure 36 increase penalties for drug and theft crimes. Yes. What this actually does is roll back Proposition 47 which passed in 2014, and reduced penalties. The real problem here is law enforcement, not penalties. ==== COUNTY (Los Angeles) District Attorney: Nathan Hochman. This is entirely a vote against incumbent George Gascon whose progressive policies have hurt Los Angeles. Judge Superior Court No 39: Steve Napolitano. A tough choice as both candidates appear qualified. Judge Superior Court No 48: Renee Rose. Again, both candidates appear qualified, Rose has more endorsements from other judges... Judge Superior Court No 97: Sharon Ransom. Picking a DA over a public defender, both seem qualified, I like Ransom's endorsements. Judge Superior Court No 135: Georgie Huerta. Both candidates are deputy DAs and appear qualified; tough call. Judge Superior Court No 135: Tracey Blount. Seems experienced and qualified. County Measure G: change to LA County governance. Yes. Makes the head of the county an elected official, expands board of supervisors. Not sure of the cost implications (supposedly budget neutral?) but seems like a reasonable restructuring. ==== DISTRICT (Los Angeles) Measure E fire protection tax. No. I support fire fighters of course but once again, do not like special purpose taxes. If this spending is needed let's put it in the District budget. ==== CITY / LOCAL Las Virgenes School District Board (2): Alan Lazar, Linda Menges. Menges is an incumbent and seemingly doing a solid job, Lazar has made STEM support a priority, which I support. LA Community College District Board of Trustees, Seat 1: Andra Hoffman. Hard to pick here but Hoffman is the incumbent and has a lot of endorsements. LA Community College District Board of Trustees, Seat 3: David Vela. Another tough pick, Vela is the incumbent. LA Community College District Board of Trustees, Seat 5: Nichelle Henderson. Henderson is the incumbent and current LACCD President (elected within the Board). LA Community College District Board of Trustees, Seat 7: Kelsey Lino. Lino is incumbent and opponent doesn't appear to have much experience. Member of State Assembly, 42nd District: Jacqui Irwin. Irwin is a conservative Democrat (not to say a conservative) running for her 5th term, seems to be doing a good job. United States Representative, 26th District: Michael Koslow. I know Michael, he's a member of the Westlake YC, and I'd like to see Brownley defeated, but seems unlikely. Weirdly, this year the ballot sorted as city/local, district, county, state, national, in the past it was reversed. Maybe this is to encourage more people to vote for the local offices? I have an unpopular opinion on this, if you don't know what you're voting for, then don't vote. If all you know is your presidential choice, then that's all you should vote for...
Of course he was not a dictator and did not do horrible things; all the weird stuff you read that might happen if he is re-elected is just propaganda. (For example, he is not in favor of an abortion ban; Roe vs Wade was a states' rights case, not an abortion ruling.) Trump has already been President and has been a public figure for decades, we know what we are going to get. You might not like it and you might not vote for him, but don't lie about what might happen.
So here we are in 2024. Biden was exposed, and Harris was chosen to take his place. She doesn't seem bright and can barely speak, and there is no reason to believe she'll reverse any of the policies which have led to our huge national debt and illegal immigration crisis. She won't be "in charge" and won't be a leader, but the same apparatus which has operated under Biden would continue under Harris. And I fear the erosion of our freedoms will continue. So yeah, I'm voting for Trump, and against Harris.
The other day I posted a pirate puzzle: find R. You could probably figure out how to proceed: Okay, so we have five equation with five unknowns. A little algebra, and we should be good. Once we know x, we can subtract 3+4+5 from x^2 and we're done. So in 2024 we could just ask an AI chatbot to do this for us, right? Well, no. I first went to ChatGPT, and it got massively wrapped around the axle, and gave me non-solutions with negative numbers and square roots and all kinds of junk. I asked it to test its answers, and it did, and then it apologized for giving me the wrong answers (!) and tried again, and again failed. Boo. So I tried Claude, and it was worse. It kept telling me how to solve the problem and asking me if I wanted it to continue, without actually solving it. When I asked it to actually keep going until it had an answer, it tried, and then informed me the problem was unsolvable. Fail. Next I tried Perplexity, which I usually use as a browser search plugin, but sadly it faired no better, giving me wrong answers, and then failing to plug the values into the original equations to test them. Finally I tried Llama. It kept getting wrong answers after some convoluted guessing. When I asked it to check, it did, admitted it was wrong, tried again, but got the same wrong answers. And also observed the problem might be unsolvable! Wow. So AI chatbots, impressive as they are, cannot do algebra? Okay, back to ChatGPT again. This time I informed it this was simple: five equations in five unknowns, just use algebra, give me positive solutions. This did the trick. The problem was it didn't know which approach to take. Impressively, in the result there was a link to the Python program it wrote and ran to get the solution. (That little blue "(>_)" at the end...) And also, the answers actually work :)
Let's check out the program, shall we? There you go. Armed with these answers - especially x = 4.67 - now we can compute R: And so R = 9.8. Does this feel right? Well 3+4+5 = 12 and R appears to be a bit less than half of the square, so yeah, it works. The most fascinating thing about this for me was the AI chatbots getting the wrong answers, and then having to coach ChatGPT into getting the right one. So much like a person. We have now *totally* blown past the Turing test!
You might be cool, but are you "Ori dressed as a Minion horizoning the fleet in the Westlake Yacht Club's Trick or Treat regatta" cool?
A photo shoot disguised as a wedding ... and an amazing day. Awesome! 🎂🤗🥂🥳
Good morning on a big day ☀️👰♀️🤵♂️🥂!
Interesting little engineering puzzle. What’s R?
EOQ! Wow, can't even believe we are at the end of another third quarter. And you know what that means: Halloween, Thanksgiving, Christmas, and New Years', bam bam bam bam. And poof another year in the books. Happens faster and faster every year. I've been ruminating on the "dwell time" of an online post. When you post something to a blog, on X, on Facebook, wherever your friends / followers / whomever see it for as long as it is your "current" post. As soon as you post something else, it gets pushed down the stack a bit, and people are less likely to see it. This is independent of the relative value or importance or length of the post. I could post a lengthy diatribe with incredible analysis, and then immediately post something trivial, and that diatribe will become history while the trivia might live on a long time.
Some blogs (and some social media) have a "keep at top" feature, which artificially boosts dwell time. I used to do this myself, every once in a while, but haven't for a while. I have mentally played with the idea of recording inbound traffic (hits) for each post, and sorting archived posts by "max hits". That would make reader interest a part of dwell time, which would be good. Presumably if something was of interest to many or externally linked it likely would be of interest to the next reader. Or I could just keep posting whatever whenever and let the chips fall where they may. Heh. {Update: added digression: whoever wanted Google's .webp images? Why do they exist? Someone spent all that time inventing a new image format just so we all have to convert them back to JPEGs. What an incredible waste of time.}
I'm still riding Centuries, and yesterday rode a nice one: the Lighthouse Century from Morro Bay up PCH to the Piedras Blancas lighthouse, through San Simeon, and by the Elephant Seal overlook. To get to 100 miles the organizers start with a nice little jaunt into the hills of Paso Robles; beautiful wine country, if you can take a minute from climbing to take in the views. Yes, it was a great ride, and yes, I am quite proud of myself for being able to ride a Century in under six hours; "only" 5,200' of climbing... but still! I've been playing with the various ways to record video of a ride from my GoPro, and this little clip contains several of them: straight video, speeded up, time lapse pictures, turned into a video, and "posterized" video to turn it into effectively a timelapse sequence. Not sure yet of the best way. I do like taking pictures during the ride and embedding them; for all that the GoPro is amazing, the iPhone pictures are better yet, and you can really see the difference on a cold and cloudy day. And with my phone I can shoot "sideways", whereas the GoPro is always pointed "down the road". Onward!
Among all the amazing capabilities of Perplexity* and honestly, c'mon, it is amazing, right? one of the biggest for me is that it has completely replaced online help. For some time now we were sure that Google was always better than a product's own online help, but now Perplexity is amazingly better.
I'm editing a video with Adobe Premiere - a timelapse of the Lighthouse Century I rode yesterday, since you ask :) - and wanted to know how to change a video segment from 29.97fps as recorded by my GoPro to 1fps as I want it added into my video. So I asked Perplexity*: "Adobe Premier how do I change the frame rate of a segment" and it tells me to use "Interpret Footage". Well that's a good answer but not the one I want, because it changes the segment's duration, so I next type** "without changing the duration" (leveraging context) and it tells me "use the Posterize Time Video Effect" which is exactly the right answer. Saved me many many minutes, and essentially makes Premiere itself more powerful, since it opens up many more of its capabilities for me.
And you know what Perplexity costs? $0... What a time to be alive!
Passwords are a problem, for sure, for me as a user as well as for me as an applications developer. Good passwords are hard to remember, every site has different rules, sometimes you have to change them, sometimes you can't reuse them, and everybody writes them down insecurely. (Yep, you do too, admit it.) So when passkeys were invented, everyone said yay. But they don't solve all the problems and create many new ones. The difficulty of having them across multiple devices, the difficulty of creating them in the first place, and the difficulty of implementing them. And the reliance on central authorities.
The best solution to passwords is not to have them at all. Just send the user a limited time link in text or email. This is simple to explain, simple to use, simple to implement. And no less secure than passwords; most of the time you can change or recover a password with a link in text or email anyway. Oh, and it supports multiple devices easily.
So long passkeys, we hardly knew ya...
Tonight I watched stage 16 of La Vuelta (the Tour of Spain cycling race), which finished atop the legendary Lagos da Covadonga climb in Asturias, the Northernmost province of Spain. It was a fantastic race on an amazing track.
I rode this very climb myself, way back in 2007; I was in Spain for business and detoured up to Asturias to ride the stage and then watch the pros do it too. After an improbable series of barriers surmounted I made it, probably the hardest climb I've ever done then or since. Watching the race today was doubly enjoyable remembering being there which seems like yesterday, 17 years ago...
Rereading my report on the day, I remember feeling a weird sense of inevitability; with each obstacle it felt like something good was going to happen, and then it did. "How did I get here (!)" indeed ... Quite a day.
Watching pro cycling in Europe is the best - in addition to another Vuelta stage I've been privileged to watch two stages of the Tour de France, and to ride and then watch the Ronde van Vlaanderen (Tour of Flanders). I still have to see a Giro d'Italia stage (Tour of Italy), and would love to see the Strade Bianca (which tours Tuscany and finishes in La Plaza in Siena). It's nice to have goals :) Onward !
a Labor Day spent sailing w your granddaughter is pretty perfect
Yesterday I drove my bike up to Arroyo Grande and took a nice out-and-back ride to Avila Beach. Here's a composite video from the ride: This is a test, there are several things going on here. First the video itself is a GoPro time-lapse combined with a Strava GPS flyby, with embedded iPhone pics. Experimental, and I like it. Composed with Adobe Premier. LMK what you think.
And LBNL, this is my first time posting a video with my new spiffy (shaved-yak) email-to-blog mechanism. I love it by the way; this is being composed on my iPad, sitting in my den, miles away from my laptop and office.
{Another Update: modified the API to support embedding videos, as opposed to merely linking them. This takes advantage of modern browsers' support for video; in the bad old days I used to embed various video plugins, and in the worse older days used Flash. But now even mobile browsers can play video directly ...} Well enough of the navel gazing; it was a fun ride watching the masses at the beach, and traversing through Shell Beach which is a beautiful little residential neighborhood tucked in between Pismo and Avila. Pismo Beach is a real slice of Americana; except for the size of the people and their tattoos, could have been any time in the 1980s; the kids and dogs playing in the surf were just the same. Speaking of dogs, this has to be the best way to tour around with them, and is my favorite pic from the trip. Woof. And onward...
I recently reread this post about AI and emergent properties, from May 2003, and was blown away by its prescience and relevance. (In posting about old posts I seem to be following the New Yorker's example and reusing content in lieu of new thinking, but in this case it's delightfully "meta", as new observations have "emerged" over time). Little did I know or could have known that 21 years later AI would be at the forefront of all tech, and that it would be a "brain dead" form of AI, without heuristics, happily using applied statistics to synthesize emergent properties. We can now hypothesize that not only does this lead to "intelligence", but it might be all that ever does; there are no underlying heuristics at all. In this, I find an analogy to Alpha Zero, which learned to play great chess (and famously, even greater Go) simply from the rules of the game, without any heuristics.
In the post linked above AI pioneer Marvin Minsky was quoted as saying "AI can't deal with concepts like water is wet'". That was true in the 1970s. And it was thought that to deal with this, "wet" would have to be defined, and an association between "water" and "wet" would have to be made. Now we can see that a concept like "wet" emerges from the presence of things like water, and so the association is causative, water "causes" wetness. Such properties are a function of observation, they are not inherent nor are they explicit, and they are to some extent influenced by the observer as well as the thing itself. Consider a more abstract property like "beauty". Not only is it famously "in the eye of the beholder", but it is only such, it does not exist in and of itself. (This point is made in another old post made only a week later, God and Beauty; I did not make the connection at the time!) Labeling things as "beautiful" does not make them so, and a definition of beauty without examples doesn't get very far. We can describe the effect of its perceived presence on an observer, and the commonality between things exhibiting "beauty" is mostly in these effects, not the described objects. (What other commonality exists, for example, between a beautiful person and a beautiful algorithm ... or a beautiful philosophy?) Fairly recent posts (well last handful, anyway):
For older posts please visit the archive.
|
|