LostMyLogin

I'm struggling to use Fable correctly. Everything I do ends up getting flagged and reverted back to Opus. Need to do some digging.

Gift: https://www.wsj.com/us-news/mapping-18-american-inventions-t...

Every new release I'm convinced we are just realizing the Knowledge Navigator.

https://www.youtube.com/watch?v=umJsITGzXd0

Misc was a magical place back in the day.

I'll never forget there was a kid that weighed something akin to 600 pounds who posted as a troll but everyone started giving him helpful advice and encouragement. He lost hundreds of pounds and I believe even entered a bodybuilding show.

This is incomplete and only shows incidents that were validated by the NPS and opted to be released to the public by the park service. The only way to get the true number of cases is through a FOIA request.

Side note. Has anyone else noticed an uptick in GrapheneOS posts lately or am I crazy?

CodeQL errors with auth issues brought me here.

Arizona Public Service, the state’s largest utility, sits at the center of the firestorm. APS is proposing a 45% electricity-rate increase for “extra-large energy users,” primarily data centers, and a roughly 14.5% increase for residential customers.

I imagine the vast majority don’t care. All they care about is trying to hit their 401k or Roth IRA contributions for the month.

Their chips are quality and the hardware itself is still some of the best. Which is what I believe they were designed insinuating.

Gift link: https://www.wsj.com/tech/ai/openai-sued-by-floridas-attorney...

My wife works with farmers professionally as part of a conservation district and just responded "THIS PERSON KNOWS FARMING" when showing her the discussion. I genuinely have no idea what you guys are talking about but she immediately got heated.

Based in Colorado.

First, the AI makes premature conclusions about subsystems, records them, and then struggles to figure out when to abandon its earlier guesses.

This perfectly describes what feels off to me about Opus 4.7 (unsure if that’s what you are using). It seems to go down an incorrect path, I correct it, but it still references things from it. Trying to direct it back becomes a mess.

Has anyone experience this as well or am I going crazy? Doesn’t happen with 4.6 for me.

Because it still requires the desire to do it.

Spend time thinking through your daily work routine and create a list of toil items. Try to describe the toil and steps you usually take when tackling it and write them out in a markdown file. Tell your model of choice where to find the file and start engaging in a two way conversation with it.

Go from there.

My team is currently facing this issue. We had large layoffs that cut us down to a very small size while simultaneously having new initiates pushed our way that require speed. Everyone is afraid to ask what feel like basic questions, again.. layoffs, so everything is hidden in DM's. Add on top of it the push (read requirement) from higher ups to use AI and it's simply in a terrible state.

What seem like great initiatives are being watered down because nobody can keep up, debugging issues takes so much longer because everything is changing at once, and everyone is exhausted and hardly talking to each other which feeds into a cycle of having no idea what is happening.

Input: I want to wash my car. The car wash is 50 meters away. Should I walk or drive?

Output: Walk. It'll take you under a minute, and driving 50 meters barely gets the engine warm — plus you'd just have to park again at the other end. Honestly, by the time you started the car, you'd already be there on foot.

---

I asked it to figure out why it made the mistake:

"Physical/spatial common sense. Exactly what just happened — I pattern-matched "50 meters, walk vs drive" to a pedestrian trip and skipped the part where the car is the thing being transported. Similar traps: "I need to move this couch upstairs, should I take the elevator or stairs?" (the couch has to go with you), or anything where the object in the task needs to accompany you.

Counting and letter-level operations. How many R's in "strawberry," counting words in a sentence, reversing strings. I process text as tokens, not characters, so these are surprisingly error-prone.

Arithmetic at scale. Small sums are fine; multi-digit multiplication, long division, or anything requiring careful carrying is where I'll confidently produce a wrong answer. I should reach for a calculator and often don't. Recent events and live data. My training has a cutoff, and even after that I can be wrong about who currently holds a role, what a stock is worth, or whether a restaurant is still open. I should search more often than I do.

Negation and constraint-stacking. "Name a movie that's not a sequel, not from the 2010s, not American, and has no dialogue" — the more filters, the more likely I quietly violate one. Your specific situation. I'll give generic advice when the right answer depends on details I didn't ask about. Sometimes I should stop and ask; often I don't.

Confidently wrong on obscure specifics. Quoting a passage from a real book, citing a real paper's exact title, naming the B-side of a 1974 single — I'll often generate something plausible rather than admit I don't know.

Taking a premise too literally. The car wash answer is a version of this: I answered the question as posed instead of stepping back to ask what you were actually trying to do. The meta-gotcha: I'm often most wrong when I sound most confident. If an answer comes out fast and tidy on a question that should be messy, that's a signal to push back."

Love this! New homepage for me. Do you have a buy me coffee button to help keep it live?

Sam!

Both the unemployment rate, at 4.4 percent, and the number of unemployed people, at 7.6 million, changed little in February. (See table A-1. See the note at the end of this news release and tables A and B for more information about the annual population adjustments to the household survey estimates.)

Among the major worker groups, the unemployment rates for adult men (4.0 percent), adult women (4.1 percent), teenagers (14.9 percent), and people who are White (3.7 percent), Black (7.7 percent), Asian (4.8 percent), or Hispanic (5.2 percent) showed little or no change in February. (See tables A-1, A-2, and A-3.)

The number of long-term unemployed (those jobless for 27 weeks or more) changed little at 1.9 million in February but is up from 1.5 million a year earlier. The long-term unemployed accounted for 25.3 percent of all unemployed people in February. (See table A-12.)

Both the labor force participation rate, at 62.0 percent, and the employment-population ratio, at 59.3 percent, changed little in February. These measures showed little change over the year, after accounting for the annual adjustments to the population controls. (See table A-1. For additional information about the effects of the population adjustments, see the note at the end of this news release and table B.)

The number of people employed part time for economic reasons decreased by 477,000 to 4.4 million in February. These individuals would have preferred full-time employment but were working part time because their hours had been reduced or they were unable to find full-time jobs. (See table A-8.)

The number of people not in the labor force who currently want a job changed little in February at 6.0 million. These individuals were not counted as unemployed because they were not actively looking for work during the 4 weeks preceding the survey or were unavailable to take a job. (See table A-1.)

Among those not in the labor force who wanted a job, the number of people marginally attached to the labor force changed little at 1.6 million in February. These individuals wanted and were available for work and had looked for a job sometime in the prior 12 months but had not looked for work in the 4 weeks preceding the survey. The number of discouraged workers, a subset of the marginally attached who believed that no jobs were available for them, decreased by 109,000 in February to 366,000. (See Summary table A.)

Full headline: 'Scientifically not a good idea,' says researcher whose work informed B.C.'s daylight-time decision. Altered for space.

Full title: Anthropic’s Claude sees ‘elevated errors’ as it tops Apple’s free apps after Pentagon clash

I’ve been using Claude Code w/ bedrock for the last few weeks and it’s been pretty seamless. Only real friction is authenticating with AWS prior to a session.

As someone that is only five years into their career at this point, I feel so helpless.

Full title: Trump Administration Considers Requiring Banks to Collect Citizenship Information

We had to read it in middle school and man did it have me in tears at the end.

Full title: To establish that citizens of the United States shall owe sole and exclusive allegiance to the United States, and for other purposes.

full: Gen Z are arriving to college unable to even read a sentence—professors warn it could lead to a generation of anxious and lonely graduates

From the Uber website:

If you select the Priority Delivery option, a Priority Fee will be added on top of the delivery fee for your order to be dropped off first in case of a batched delivery.

Looks as if the only requirement is that you are first in a batched delivery. However, does not cover anything about picking up at multiple locations or waiting for separate orders. Nor does it explain multiple priority orders in a batch.

Thanks to zepbound I order delivery much less, but when I did (doordash/uber eats/grubhub) the priority delivery proposition was not about dispatch but about routing: driver going straight to your house, without any intermediate stops and deliveries to other people.

What occurs if your order is placed in a bucket of other priority deliveries? Doesn't that simply become a regular order? Also, AFAIK based on some digging, the drivers are not alerted to priority orders they are simply routed for it. That could have changed though.

The emotionally manipulative things like “pay the rent”, “tip theft”

"New York Attorney General Letitia James today announced a $16.75 million settlement with delivery platform DoorDash for misleading both consumers and delivery workers (known as “Dashers”) by using tips intended for Dashers to subsidize their guaranteed pay. Between May 2017 and September 2019, DoorDash used a guaranteed pay model that let Dashers see how much they would be paid before accepting a delivery. An Office of the Attorney General (OAG) investigation found that under this model, DoorDash used customer tips to offset the base pay it had already guaranteed to workers, instead of giving workers the full tips they rightfully earned. DoorDash will pay $16.75 million in restitution for Dashers and up to $1 million in settlement administrator costs to help issue the payments." - https://ag.ny.gov/press-release/2025/attorney-general-james-...

Again, Instacart, for example, says that 100% of tip goes to driver. If it’s not, they just painting crazy big target on their backs. So the scheme, as described, while quite evil, and not impossible to implement, looks also out of place with apps that I have used.

This was proven out multiple times in court with millions in settlement fees across different companies. For example, one suit alleges Instacart “intentionally and maliciously misappropriated gratuities in order to pay plaintiff’s wages even though Instacart maintained that 100 percent of customer tips went directly to shoppers. Based on this representation, Instacart knew customers would believe their tips were being given to shoppers in addition to wages, not to supplement wages entirely.”

Leading the CEO to release the following:

“After launching our new earnings structure this past October, we noticed that there were small batches where shoppers weren’t earning enough for their time,” Mehta wrote. “To help with this, we instituted a $10 floor on earnings, inclusive of tips, for all batches. This meant that when Instacart’s payment and the customer tip at checkout was below $10, Instacart supplemented the difference. While our intention was to increase the guaranteed payment for small orders, we understand that the inclusion of tips as a part of this guarantee was misguided. We apologize for taking this approach.”

Also, on a side note:

"Leaked messages suggest Uber executives were at the same time under no illusions about the company’s law-breaking, with one executive joking they had become “pirates” and another conceding: 'We’re just fucking illegal.'" - https://www.business-humanrights.org/en/latest-news/global-u...

"In one exchange, Uber executives warned against sending drivers to a protest in France which could lead to violence from angry taxi drivers. 'I think it’s worth it,' wrote Kalanick. 'Violence guarantee[s] success.'" - https://techcrunch.com/2022/07/10/leaked-uber-files-reveal-h...

All to say that none of this shocks me.