Random Bits

* Today, I learned that there is a genre called "HFY" which means "Humanity, Fuck Yeah!", that flips the trope of advanced and superior aliens to portray humans an exceptional and lauds our species' positive traits against the space opera foil of other more established interstellar aliens types who lack those traits. I'd read stories in this genre many times before but never realized that it was a genre.

* A psychology tiktok discussed how the easier learning stays the less it sticks in the long run. It recommends focusing on testing yourself, closed book, on what you can recall, spacing study sessions weeks apart, intertwining concepts and types of problems rather than studying them in blocks of the same thing, restating concepts you've studied in your own words, and focusing on connecting new facts to existing frameworks and related concept. It cites:

Bjork, R.A. — "Desirable difficulties" framework 

Bjork, R.A. & Bjork, E.L. (1992) — New theory of disuse 

Roediger, H.L. & Karpicke, J.D. (2006) — Testing effect study

Karpicke, J.D. & Roediger, H.L. (2008), Science — Retrieval practice study 

Pashler, H., McDaniel, M., Rohrer, D., & Bjork, R. (2008) — Learning styles literature review 

Rohrer, D. et al. — Interleaved vs. blocked practice studies 

Chase, W.G. & Simon, H.A. (1973) — Chess expertise study

* In the not new, but still true department:

* One of the truly striking things about the 2026 election is how high the quality of the Democratic candidates is and how low the quality of the Republican candidates is. The Democratic candidates are moderately better than average. The Republican candidates are truly pathetic. As a New York Times opinion column explains:

In race after race, Democrats have managed to recruit and advance candidates who are well positioned to capitalize on changing political circumstances. Even if he were not facing off against Ken Paxton, an experienced but still strikingly weak candidate, James Talarico would still be an impressive recruit. He is a natural communicator — whose seminarian background and personal faith have been a great asset in winning over Texas voters — and he has proved to be both nimble and aggressive, able to hold Paxton’s feet to the fire despite facing an onslaught of cash from outside groups and actors, like Elon Musk.

You could say similar things of Adam Hamilton in Kansas, perhaps the surprise candidate of this cycle, who is running a competitive race against the Republican incumbent, Roger Marshall. Hamilton is a religious leader — the founder and senior pastor of the Church of the Resurrection, the largest United Methodist congregation in the United States — with broad appeal in a state that is, like Texas, known for its deep religious feeling. With his moderate politics, Hamilton has successfully presented himself as a bipartisan figure, with support from prominent Republicans in Kansas. It also helps that Marshall, a doctor, has struggled in the face of revelations that he took harsh and punitive action against his patients, dragging them to court for trivial amounts of money.

Other Democratic candidates, such as Josh Turek in Iowa, Sherrod Brown in Ohio and Troy Jackson in Maine, have also successfully capitalized on the Republican Party’s worsening electoral position. And even those Democrats who might be a little more progressive than their electorates — Peggy Flanagan in Minnesota and Abdul El-Sayed in Michigan — show a great deal of political talent and are well-positioned to win.

As for the Republicans, the opposite appears to be true. There are a few candidates in the field who may well outperform the fundamentals, like the Iowa Senate nominee Ashley Hinson. Michele Tafoya in Minnesota also seems to be keeping up with her opponent. But beyond them, Republicans are working with candidates who are just not up to snuff. In Georgia, Mike Collins has been marred by his associations with white nationalists and other extremists, and is headed for a crushing defeat in his race against the incumbent, Senator Jon Ossoff. Darline Graham in South Carolina is barely ahead of her Democratic opponent, and again, this is South Carolina. The aforementioned Ken Paxton is such a lackluster candidate that he may well deliver the Texas Republican Party its first statewide defeat for federal office since 1988.

Perhaps the most striking example of a candidate ill-equipped for a competitive race is Vivek Ramaswamy, the Republican nominee for governor of Ohio. Ramaswamy ran unsuccessfully for the 2024 Republican presidential nomination, then returned home to Ohio to try to lead the party there. Perhaps the fact that Republicans were not anticipating a competitive year is how Ramaswamy managed to win the nomination. Either way, he is currently poised to lose to the Democratic nominee, Amy Acton.

In apparent desperation, the Ramaswamy campaign commissioned an advertisement featuring Amy Acton’s mother. The two have not seen or spoken to each other since 1978. As a child, Acton testified against her stepfather, who was charged with sexually abusing her. Acton’s mother took her husband’s side, and the two fled the state. In the advertisement, Acton’s mother calls her daughter a liar, someone who “lies about her past.” The evidence says otherwise. To give one example, documents released by Mahoning County officials in Ohio corroborate Acton’s claim that she lived in a tent as a child.

The Ramaswamy camp, which includes fellow Ohioan JD Vance, thinks the ad is effective. But it’s been condemned by people on both sides, including the state’s Republican governor, as crude, craven and cruel. There is a chance, as well, that it blows back on Ramaswamy, undermining his campaign in the critical final stretch of the election. It is the kind of move that a more seasoned and competent candidate might have avoided. As it is, Ramaswamy’s gambit may well drag down the entire statewide ticket. 

* Then there is this (extracted from this Facebook reel):

* The U.S. Supreme Court has really, really failed us in these last couple of years. They were the part of the system that was supposed to prevent so many of the abuses of Trump 2.0. And, the lower trial and intermediate appellate courts, on balance, did much better.

* Is our system broken? It certainly isn't perfect. But the midterms, if the Democrats make an unprecedented surge to gain control of both houses of Congress, and those elections are honored, and Democrats in control of Congress manage to exert significant power, could show that it isn't completely broken. The basic design feature of a democracy is that is the party in power screws up in an unpopular way, that voters will replace them and deprive them of power. Voters get to do so every two years in the U.S., which is a shorter reaction time than in parliamentary systems where a prime minister can lose the confidence of parliament on very short notice and lose power, but the U.S. makes up for it a little by having strong federalism and powerful courts to slow down serious misconduct and bad policy at the top. If we survive this crisis for now, does that show that our system works? Or are we just hitting a temporary respite from a decline driven by a broken system?

* Another feature of a Democratic party majority in Congress, is that it changes the logic of political succession. It puts a Democrat right after the Vice President in the line of succession. If Trump, who is not a very healthy man and 80 years old, were to die or be so disabled that he couldn't function, or was removed via the 25th Amendment, then J.D. Vance would be President, and J.D. Vance couldn't fill the Vice Presidency without cooperation from a Democratic Party controlled Congress. And, in the absence of a Vice President, the Democrat who is elected to be Speaker of the House would be next in line for the Presidency. While most Presidential assassinations aren't attempted by rational people (although there have been three apparent Presidential assassination attempts so far, one of which can close to being successful, so far in Trump 2.0), rationally, it doesn't make a lot of sense to try to assassinate a President when the next dozen people in the line of succession are just as bad. But, if either J.D. Vance or Trump ceases to hold their current office, then the nation is just one assassination away from a complete flip of control of the Executive branch to a politically dramatically different alternative, a Presidential assassination makes a rational political sense at a fairly narrow immediate level, and the list of people or organizations with an incentive to do it then (many of them powerful and well-organized), gets longer than the line at the DMV. To be clear, I'm not advocating that anyone assassinate anyone (and an assassination isn't even necessary, just circumstances that cause two elected officials to vacate their offices). Undermining the norm against assassinating politicians (which is stronger in the U.S. than in many countries), opens up a huge can of worms that poses an immense risk of ending democracy in the medium term, even if it stops a despot in the short term. I'm just observing that the political logic for someone to do so changes dramatically. And, since the system that was supposed to deal with this situation in the constitution (impeachment) has proven to be completely inadequate (since it requires a two-thirds majority support in the U.S. Senate which has shown itself to be entirely partisan in its behavior, and a two-thirds Senate majority is almost impossible to achieve), there is strong pressure from the system to find other ways to remove grossly of the rails officials.

* Prediction markets are tolerably accurate at predicting things in many domains.

* What would a society where inheritance and nepotism in business and wealth succession were largely gone, in analogy to ending monarchy and aristocracy, look like? One recent study in historical economics suggests that the impact wouldn't be all that large, but it has limitations that make me doubt it (it is based largely on the impact on inheritance amount and likelihood based upon random variations in family size).

The 5th Circuit En Banc Decision On Pre-Hearing Immigration Detention

One of the key issues in immigration law is whether someone detained for being an undocumented immigrant pending a removal hearing in which that person has statutory grounds upon which they could not be deported, who is detained in the interior of the U.S. rather than at the border, and has long standing ties to the U.S. as a resident of the U.S. is entitled to a bond hearing that evaluates their dangerousness and flight risk.

The second Trump administration has taken the position that an undocumented immigrant who awaits a removal hearing may be detained indefinitely, no matter how long the hearing is delayed, no matter where they were detained or how long they have been in the U.S. or what grounds they have to not be deported, and that these individuals have no right to a bond hearing to give those individuals an opportunity to be released before their removal hearing. (The Trump Administration has also explicitly told private contractors operating immigration detention centers that the conditions of confinement should be cruel, but that is a story for another day.)

Thirty years of prior practice of the Executive Branch, prior to Trump 2.0, under the immigration statute adopted in 1996 and since amended in ways largely immaterial to this question, have held that there is a right to a bond hearing to assess whether someone awaiting a removal hearing is dangerous or a flight risk, while that removal hearing is pending (and at a time when the government has not established before an immigration judge that they are entitled to deport someone). This was also the prevailing practice pre-1996, including in the pre-1907 period which the majority mischaracterizes, as a dissenting opinion notes. As explained at page 136 of the Slip Opinion in a dissenting opinion:

The Government and the majority repeatedly cite the Supreme Court’s footnote in Demore v. Kim, 538 U.S. 510 (2003), that “prior to 1907 there was no provision permitting bail for any aliens during the pendency of their deportation proceedings.”  Id. at 523 n.7.  That sounds significant, but, prior to 1907 this country simply did not detain persons potentially unlawfully present in the interior except in extraordinary circumstances and under narrowly drafted laws; and, when it did so, the general rule was that bail would be available.  See Nash, supra, at 1540–41, 1564–76.4  In other words, the practice of seeking out people unlawfully present in this country in the interior en masse is an invention of modern immigration law.  The practice (challenged here) of seeking out persons unlawfully present in this country in the interior and subjecting them to detention without meaningful process or the possibility of bail—that is, to what was long understood to be the statute applicable at ports of entry, § 1225(b)(2)(A), rather than the interior detention statute, § 1226(c)—is a recent invention of the Department of Homeland Security in the last year or so contradicting its almost three decades old practice.   

Nine of the eleven numbered geographic circuits of the U.S. Courts of Appeal (ignoring the U.S. Court of Appeals for the Federal Circuit which doesn't have jurisdiction over these cases) have held that there is a right to a bond hearing under the applicable U.S. Supreme Court precedents (admittedly, most of which were decided before the current ultra-conservative majority arose in the U.S. Supreme Court, and the current U.S. Supreme Court often shows little regard for its prior precedents or a lower court near consensus). There does not appear to be a case on point out of the U.S. Court of Appeals for the D.C. Circuit.

So did a 2-1 decision of a three judge panel of the 5th Circuit, and three U.S. District Courts who ruled in favor of the people seeking bond hearings that the government appealed to the 5th Circuit. So did 7 of the 17 judges of the 5th Circuit in a sua sponte review en banc in an 159 page set of majority, concurring, and dissenting opinions announced today of that 2-1 decision in the case of Rodriguez v. Vergara (and two other cases consolidated with it).  

Just counting noses in the 5th Circuit who have considered this case (including the trial court judges in these cases), 10 judges ruled that there is a right to a bond hearing and 10 judges ruled that that there was not. It also bears noting that even the current extremely conservative U.S. Supreme Court reverses decisions of the 5th Circuit more than any other U.S. Circuit, generally, in a less conservative direction than its decisions. 

The 5th Circuit en banc majority relies heavily on a U.S. Supreme Court case involved sex offender registration for a general principle (a circumstance where dangerousness is a presumptively present in almost every case), which is Conn. Dep’t of Pub. Safety v. Doe, 538 U.S. 1, 8 (2003), rather than prior U.S. Supreme Court immigration cases on point, in a very strained analogy.

The only other U.S. Court of Appeals denying bond hearings is that 8th Circuit (which is also very conservative, although so is the 11th Circuit that went the other way), and the 8th Circuit which also followed a related 5th Circuit precedent, Buenrostro-Mendez v. Bondi, 166 F.4th 494 (5th Cir. 2026), cert. sought and neither granted nor denied at this point, in which the 5th Circuit make a minority interpretation of the relevant immigration statute. See Slip Op. at pg. 100 (in a dissenting opinion). See also Slip. Op. at pg. 141 (in a dissenting opinion) ("On appeal, only this court in Buenrostro-Mendez v. Bondi (and the Eighth Circuit, following us) has accepted the Executive Branch’s statutory arguments for its recent about-face.  See 166 F.4th 494, 502–05 (5th Cir. 2026); Avila v. Bondi, 170 F.4th 1128, 1132–36 (8th Cir. 2026).").

District judges reportedly have rejected the government's position in more than 9,300 rulings (see also page 154 of the slip opinion), relying especially upon Zadvydas v. Davis, 533 U.S. 678 (2001), and three certiorari petitions are pending. As noted at page 144 of the Slip Opinion: "The Supreme Court thus far has not accepted certiorari in the sole petition to squarely pose the due process question.  See Petition for Writ of Certiorari, Putra v. LopezCampos, No. 25-1415 (U.S. filed June 22, 2026), https://perma.cc/8TXG-9S4J.  Rather, it first will assess the full statutory picture. See Rhoney v. Barbosa da Cunha, No. 26-104, 2026 WL 2942994 (U.S. Oct. 1, 2026) (granting certiorari)." 

Another key recent U.S. Supreme Court decision, which the en banc majority in the 5th Circuit today and the dissenters in the same decision disagree upon the meaning of is DHS v. Thuraissigiam, 591 U.S. 103 (2020), in which the U.S. Supreme Court ruled against someone who was just 25 feet across the U.S. border regarding their immigration law rights. The majority pretty clearly misreads this decision as applying to immigrants who have lived in the U.S. for more than a decade and have strong ties to U.S. citizen children, which the decisions plain language seems to contradict.

A footnote on page 133 of the slip opinion linked above, to a dissenting opinion, sums up the circuit split:

The majority’s holding places our court in a minority position among the circuits. See Guerrero Orellana v. Moniz, 188 F.4th 1, 29 (1st Cir. 2026); Barbosa da Cunha v. Freden, 175 F.4th 61, 96 (2d Cir. 2026), cert. granted sub nom. Rhoney v. Barbosa da Cunha, No. 26104, 2026 WL 2942994 (U.S. Oct. 1, 2026); Buele Morocho v. Warden Phila. FDC, No. 261150, 2026 WL 2546223, at *10 (3d Cir. Aug. 28, 2026); Lopez Garcia v. Guadian, No. 257044, 2026 WL 2661092, at *21 (4th Cir. Sept. 10, 2026); Lopez-Campos v. Raycraft, 175 F.4th 713, 735 (6th Cir. 2026); Cirrus Rojas v. Olson, 183 F.4th 909, 931 (7th Cir. 2026); Padilla v. United States Immigr. & Customs Enf’t, No. 24-2801, 2026 WL 2822044, at *19 (9th Cir. Sept. 21, 2026); Santillan Quiroz v. Mullin, 180 F.4th 1226, 1251 (10th Cir. 2026); Hernandez Alvarez v. Warden, Fed. Det. Ctr. Mia., 175 F.4th 1258, 1285 (11th Cir. 2026). 

Without today's en banc ruling, the 8th Circuit's ruling in Banyee v. Garland, 115 F.4th 928, 933–34 (8th Cir. 2024) (holding that a lawful resident alien could be detained pending removal for over a year as long as “deportation remain[ed] a possibility”) would have been a solitary outlier.

The circuit by circuit analysis is a little complicated, because all immigrations cases funnel up to the national Board of Immigration Appeals (BIA), which like the courts it reviews is an Article II court of the executive branch that is not judicially independent, and then from there to the U.S. Court of Appeals Circuit in which the case arose at the immigration trial court level. As explained at page 154 of the Slip Opinion:

Article II IJs generally are not bound by Article III district court precedent, just that of courts of appeal.  The BIA has stated that it, “as well as Immigration Judges, [is] bound to follow [BIA] precedent, the Attorney General, and the circuit court of appeals with jurisdiction over the geographic region where a case occurs.”  Matter of Garcia, 28 I. & N. Dec. 693, 695 (B.I.A. 2023); see Matter of Anselmo, 20 I. & N. Dec. 25, 31 (B.I.A. 1989) (noting that the BIA has permitted intercircuit nonacquiescence, but “has not followed a practice of nonacquiescence in decisions of a circuit court in cases arising within the jurisdiction of that circuit”).   

But, habeas corpus review starts in a U.S. District Court (an Article III independent trial court) in a collateral attack on immigration court decisions and proceeds in the usual way to the relevant U.S. Court of Appeals Circuit.

Today's 5th Circuit en banc decision is in my view, clearly wrong under existing precedents, and affects many hundreds of thousands of people in a highly material way. There is little doubt that the U.S. Supreme Court will eventually chime in, and it is not at all clear how it will resolve the issue, since it involves a signature position of the Trump administration which the current U.S. Supreme Court has upheld in all but the most extreme instances of overreach.