edbaskerville

Interestingly, if you have one-dimensional observations f(t) of a k-dimensional strange attractor, the lagged vector time series [f(t); f(t - tau); f(t - 2 * tau); ...; f(t - (k - 1) * tau)] maps onto the full k-dimensional attractor. Specifically (as I check Wikipedia) it's a diffeomorphism, an isomorphism of differentiable manifolds.

Presumably the earth system isn't at anything resembling an attractor right now, but I wouldn't be surprised if people are trying to use related techniques to try to detect qualitative changes in the system dynamics (like bifurcations).

Maybe someone more knowledgeable could chime in on whether/how measurements at a single point on the earth's surface might be used to do that?

https://en.wikipedia.org/wiki/Takens%27s_theorem

Does anyone have good estimates of what percent of real production code is currently being written by LLMs? (& presumably this is rather different for your typical SaaS backend vs. frontend vs. device drivers vs. kernel schedulers...)

Deep cut

Solar and wind, with battery storage, can get you to say 90%, and then you only need 10% from other sources like geothermal and nuclear to fully decarbonize.

& if they live in a cool place, they're getting a small space heater as a bonus.

Can you give more details?

It wouldn't particularly surprise me if Sam Altman were racist, but I'm curious what the specific incident you observed was.

The good news is that the Trump regime is unpopular, and doing crazy things is making them more unpopular.

The bad news is they keep doing crazy things.

Political donation spam is a plague. I ran a donation website in 2018 and 2020 that split up money among many candidates, and by far the biggest complaint was the flood of email that came after using my site. In 2018 there wasn't even an opt-out button on ActBlue. In 2020 they added one, but the default was still to share your info. But it doesn't even really matter, because campaigns continue to buy and sell donor lists, so once you're in the system, you'll never get out.

It's a legal problem, in that spam laws simply don't apply to political campaigns.

But fundamentally it's a collective action problem. Excessive fundraising messages hurt the overall brand of the party and politicians in general, but for each individual politician, the advice from consultants is that each extra message has marginal value. This is actually true for out-of-district messages—they might get your money, but if they piss you off, they still don't lose your vote.

There is some movement to try to fix this.

Oath (oath.vote) is an ActBlue alternative that doesn't share your phone number or email address with campaigns. They can't erase you from the system, but at least they're trying to do the right thing.

Eventually, if groups like Oath, Crooked, Emily's List, etc. can all team up and say, hey, you won't get donations through us if you keep spamming people, we might see some change.

I assume things are also bad on the Republican side. It would be easy to say it's good if their brand suffers—but actually, I want them to start behaving more responsibly, including in this area.

Mountains Beyond Mountains was an incredible recruiting tool for health equity work, inspiring a huge number of people (including my partner) to try to follow in Paul Farmer's footsteps.

(Farmer himself died a few years ago, at only 62, of a sudden heart attack in his sleep, but he seems to have put in about 100 lifetimes worth of work. One wonders if his legendary overwork contributed to an early death.)

I'm really starting to hope the costs are unsustainable and semiconductors hit a brick wall. But I'm not optimistic.

Switched to this from Apple a year and a half ago. Works for most things. Unexpectedly, replacement apps lack polish. Also, RCS works very inconsistently (been without it for months), seems to be Google's fault. There may be workarounds, but I haven't had the energy to try the more complicated suggestions.

I am probably going to switch back to a used old iPhone for "phone appliance" tasks, but keep around the Pixel for other things.

My main takeaway from the experience is that iMessage is an even bigger weapon than I thought.

Old Macs in the background. Electronic soundtrack reminscent of Chariots of Fire, which played during the Mac intro.

Worse is better. Because better, it turns out, is often much, much worse.

It is possible for a devoted individual to do this, but it is not possible to solve deep societal problems one devoted individual at a time. We still need massive regulation of addiction-based business models.

Drove through SF this evening. Most people treated it as a four-way stop! I was generally impressed.

TikTok (along with the other platforms) is more like cigarettes, or sugar.

It's highly addictive. The negative effects are somewhat diffuse and may take a while to really impact your life, but they're very real.

And, rather importantly, it's legal and widely available, and the industry behind them is suppressing evidence of their harms and making tons of money off of addiction.

Other folks have given general answers, but I'm wondering, what ISP do you have, and where?

(I'm lucky to have Sonic, in the SF Bay Area. A local ISP that actively campaigned for net neutrality and has 1Gps symmetric as the standard basic fiber plan. Pretty sure they're not shutting down anybody's servers.)

Clever! But I think you might have forgotten to simulate moderation/dang. Sex and Nazis on there already.

In the US (edit: and elsewhere!), "beaming" worked great between Apple Newton devices, including the pretty cool eMate 300 (an early Jony Ive creation, I just found on Wikipedia).

In 1993.

> As an outsider looking in I'm starting to wonder if this world is just a bunch of academically capable but socially stunted individuals being preyed on by socially voracious people like Epstein with no morals?

The present-day tech world seems like a pretty extreme version of this phenomenon. Many of our sociopaths (e.g., Musk, Zuckerberg) got a boost from actual technical abilities along the way, which I suppose is similar to Epstein—he seems to have been pretty talented at finance.

(Edit: Musk and Zuckerberg are not socially talented in the usual sense, but have still been extremely successful at getting other people to do what they want.)

Oh man. It's not just Apple. I've had months of RCS not working on GrapheneOS, and have no idea who to blame. The first time it stopped working, I fixed it by switching carriers (AT&T -> T-Mobile). Maybe I'll try switching back! Or maybe I'll switch back to an iPhone and give in to iMessage. :(

Indeed. In the not-too-distant future where renewables are the vast majority of generation (sooner in China than in the U.S. at current rates of progress), the weather matters more and more.

You can blame Chuck Schumer, but I agree that it's wrong to blame the rest of them.

Bill Kristol has come a long way! He would vote for Mamdani...over Cuomo and Sliwa anyway.

https://www.cmcforum.com/post/bill-kristol-says-he-would-vot...

You can hear him discussing it here:

https://www.thebulwark.com/p/bill-kristol-fake-news-on-60-mi...

I'm also reminded of the time Jon Stewart got Bill Kristol to admit that a government-run health care system (the VA) was good:

https://youtu.be/rRSZiWwiBuE

They were! Typing...in Haskell...on a keyboard...on their screen...

To summarize: the ear does not do a Fourier transform, but it does do a time-localized frequency-domain transform akin to wavelets (specifically, intermediate between wavelet and Gabor transforms). It does this because the sounds processed by the ear are often localized in time.

The article also describes a theory that human speech evolved to occupy an unoccupied space in frequency vs. envelope duration space. It makes no explicit connection between that fact and the type of transform the ear does—but one would suspect that the specific characteristics of the human cochlea might be tuned to human speech while still being able to process environmental and animal sounds sufficiently well.

A more complicated hypothesis off the top of my head: the location of human speech in frequency/envelope is a tradeoff between (1) occupying an unfilled niche in sound space; (2) optimal information density taking brain processing speed into account; and (3) evolutionary constraints on physiology of sound production and hearing.

> One thing to keep in mind when using the connection pool: the connection pool needs the ability to track which connections are currently being used. To do this, the connection pool does not return Connection but ConnectionRef, a struct that derefs to Connection but contains some additional lifetime tracking.

> But Connection is Clone, so in principle there is nothing stopping you from cloning the wrapped connection and losing the lifetime tracking. Don't do this. If you work with connections from the pool, you should pass around either a ConnectionRef or a &Connection to make sure the underlying ConnectionRef stays alive.

Hmmm...

I'd like to see the incovenient API. Or maybe there's a bit more work that could be done to make it convenient? Is there an insurmountable problem that prevents completely hiding the underlying Connection?

I think the really interesting problem in this space is designing UIs and data structures that, on the one hand, capture as much user intent as possible, but, more importantly, make it easier for users to manage conflicts.

I.e., if there's a tricky conflict, the app need not resolve it at all. Rather, it should provide, by default, a nice way for the user to manage the resolution as part of the normal workflow.

Or, phrased another way, conflicts aren't conflicts. Parallel, "conflicting" edits are simply a state of affairs that is inherent to the process, and are still reflected in the data structure after merging all edits.

How this would actually look would probably vary from domain to domain. But my general philosophy on this stuff is that if complexity is real and potentially important to the user, the software should expose the complexity and enable the user to manage it, not force a simplification that hides something important.

Robert Dinero?

For a second I thought you were citing some special mythological timeline from AI folks.

Then I got it. :) Something so mundane that maybe the AIs can help prevent it.