HNHacker News
TopNewBestAskShowJobs

MarkusQ

1,979 karma · joined October 24, 2013

submissionscomments
MarkusQ··on Fed hikes rates as inflation worries push up bond yields
Depends on your time frame for "relative to"; rates have been high compared to ZIRP-era, but still moderate to low when compared to historical norms.
MarkusQ··on Fed hikes rates as inflation worries push up bond yields
> the Fed has no mechanism to really deal with supply-shock driven inflation.

Inflation is just a change in the ratio of money to stuff. You can reduce inflation by increasing the stuff or reducing the money, and cause it by doing the opposite. There's no requirement that the solution is applied to "the same factor", either works.

Sure, if you're wanting to assign blame or worried about externalities these things start to matter. But monetary policy is a perfectly fine tool for dealing with inflation, regardless of the cause.

Note: inflation causes prices to rise, but that doesn't mean that all changes in price are caused by inflation.

MarkusQ··on How much oil-market buffer is left?
Russia hasn't been exporting diesel to anyone for a while now, and they're unlikely to resume exports anytime soon.

https://www.reuters.com/business/energy/russia-extends-diese...

MarkusQ··on There's a 100% Chance AI Agents Are Ruining the Internet
> ...or whether you believe that AI continues to be a stochastic parrot, smoke and mirror plagiarism machine...

I feel heard!

What constantly amazes me is how many people don't get this after interacting with them for a while.

MarkusQ··on An atlas of periodic solutions to the three-body problem
Any set of points (w. velocities) along the trajectories could be the "starting positions".
MarkusQ··on Data collected by cars and sold to third parties
Yeah, but I still have a soft spot for the old classics like privacy, autonomy and dignity.

I don't want to give up my modern conveniences, but I have and will if forced to choose.

MarkusQ··on Julia 1.13 highlights
I really like Julia, but I wind up not using it as much as I might otherwise because the startup time kills it for many use cases (though of course it's easily amortized in others).
MarkusQ··on Data collected by cars and sold to third parties
When I was a lad we walked out to the car and did that with the key. In the snow. It wasn't that bad and, despite what you may have heard, was not in fact uphill both ways.
MarkusQ··on Navier-Stokes Announcement
More likely it's because a million dollars one way or the other won't make a difference in the hole they are digging / mountain they are building.
MarkusQ··on LRU is harder to beat than the KV-cache papers suggest
And then subtly misuse them.

LLMs write like some people cook, adding bespoke artisanal Belgian sea salt, truffle oil and weirdly specific cheese without any thought given to how the result will taste.

MarkusQ··on OpenAI’s Navier-Stokes release included a Lean 4 formal proof
You may not like the truth, but it's still the truth.

https://marketwise.com/investing/openai-losses-surge-to-21-b...

Anthropic is "profitable"... if you exclude compensation and compute cost commitments:

https://aitoolsrecap.com/Blog/anthropic-first-profit-2026-re...

MarkusQ··on Recursion into madness
Right, but if you ask it to make a systematic change (the room decor for light to dark and back) or something with subtle implications that would require other changes to fit in?

Changing the title in the movie example is this sort of change; the "change a single word" isn't analogous.

MarkusQ··on Recursion into madness
It would have been interesting to try something closer to the movie poster example; say have the AI rewrite the passage with the room toggling between the initial light-mode decor and a dark-mode decor, then back, repeatedly. Or even change the woman's name to "Mrs. Smythe" and back.
MarkusQ··on OpenAI’s Navier-Stokes release included a Lean 4 formal proof
The point was that a textbook (where the 40hr/page estimate comes from) is cumulative/linear -- what you need for page n was defined / established on the preceding pages. But in a proof such as this you can call on any other published result (and those can do the same) so the dependency graph is (potentially) much bushier. Thus later pages of the proof should take far more than 40 hours to manually formalize.
MarkusQ··on OpenAI’s Navier-Stokes release included a Lean 4 formal proof
This.

The point of these problems is the understanding / tooling gained in solving them. We're getting none of that. At best they are like a modern oracles, correctly answering your questions in a way that's doesn't help you any. (At worst,...)

MarkusQ··on OpenAI’s Navier-Stokes release included a Lean 4 formal proof
Given that they all the bit AI players are still loosing money, it follows that their total costs are _higher_ that their API pricing would imply.
MarkusQ··on Douglas Hofstadter: Analogy as the Core of Cognition [video]
By:

"an LLM AI just solved Navier Stokes"

I assume you mean:

"an LLM AI [company] just [claimed that a team of mathematicians they hired, using their AI] [may have] solved [part of] Navier Stokes[, definitely prompted by (and possibly by looking at) the work of human mathematicians."

MarkusQ··on More questions about whether researchers can trust OpenAI with unpublished math
PaaS: an acronym for "Plagiarism as a Service" which replaced the older terms AGI, GPT and LLM in late 2026. Origin uncertain.

Pass it on.

MarkusQ··on How An AI math breakthrough ignited a controversy
I think he's well on his way. Math friends that would have said "Altman who?" a week ago are now saying "Don't talk to me about that #@$!@&!"
MarkusQ··on Watch Los Angeles get built, one building at a time (1880–2026)
It's exactly the right measure to use if the claim is that they "slashed spending". They didn't. They continued increase may or may not have kept pace with other measures (GDP, population, whatever). But they didn't slash spending.

As it turns out, even your new goal posts fail the test of facts. In the years immediately following prop 13, the GDP grew at an APR < 15% [1] while the budget grew and an APR > 15% [2]. The per capita spending did slump briefly in the early 1980s (as shown in the graph on this article[3]) but that was more of a reversion-to-the-trend (which continued upwards thereafter.

While there may be some remaining tortured reading of the data under which you could claim they "slashed spending" it would be manifestly disingenuous to do so. At best the rate of increase slowed briefly.

[1] https://files.ceqanet.lci.ca.gov/276991-2/attachment/cSe45La...

[2] https://sbud.senate.ca.gov/sites/sbud.senate.ca.gov/files/Ca...

[3] https://www.mercurynews.com/2023/12/24/whats-behind-californ...

MarkusQ··on Icy Moons Are Ocean Worlds
If so, they passed up the opportunity to make another dumb Ohio joke:

"Ohio isn't actually square, though you wouldn't know that from talking to the inhabitants" or some such.

MarkusQ··on Icy Moons Are Ocean Worlds
https://everydayastronaut.com/mars-climate-orbiter/

Although, as it turns out, "orbiter" was a poor choice of name.

MarkusQ··on Watch Los Angeles get built, one building at a time (1880–2026)
They did not slash spending. At most the rate of increase slowed down for a year or so before continuing to rise at an even faster rate.

https://dof.ca.gov/about-us/history/state-budgets/

MarkusQ··on Show HN: Interactive Tree of Life
Took me a minute to figure out why I got a picture of a tree[1] for haloarchaea.

https://www.inaturalist.org/photos/660443931

MarkusQ··on Scientists observe Einstein's gravity in the quantum world
They didn't start the fire. Quantum + biology → Woo has been with us since the 1960s at least.
MarkusQ··on Scientists observe Einstein's gravity in the quantum world
Vlatko Vedral (of Oxford) to be clear.

(He's listed as V. Vedral on the Tardigrade paper you linked, and it's more common to use last names in this context.)

MarkusQ··on Watch Los Angeles get built, one building at a time (1880–2026)
That makes it sound like the increase was due to some inexorable natural force. California's per capita budget is over twice Arizona's, and they're running a huge deficit on top of that. If taxes are too high, cut taxes. If that means you need to cut spending, cut spending. But don't play "hey, let's tax the other guy!" games and expect that to work.
MarkusQ··on Every novel is boring until it isn't
"Many years later, as he faced the firing squad, Colonel Aureliano Buendía was to remember that distant afternoon when his father took him to discover ice.

At that time Macondo was a village of twenty adobe houses, built on the bank of a river of clear water that ran along a bed of polished stones, which were white and enormous, like prehistoric eggs. The world was so recent that many things lacked names, and in order to indicate them it was necessary to point. Every year during the month of March a family of ragged gypsies would set up their tents near the village, and with a great uproar of pipes and kettledrums they would..."

MarkusQ··on How accurate have Ed Zitron's AI skeptic predictions been?
They aren't synonyms and are, in general, not interchangeable. If I'm trying to get out of an unpleasant task by intentionally failing an aptitude test, the "correct" answer is going to be the least effective. Likewise if I'm crafting a click-bait headline or trying to make use of Cunningham's Law or... well, I hope you get the point.
MarkusQ··on 216M Spy TVs – The LG Smart TV Problem [video]
Whoa!

I think maybe I stepped on somebody's agenda?

Both capitalism and evolution depend on the fact that over time, on average, less fit organisms/organizations are displaced by better alternatives. You may not like that, but down-voting anyone who points it out doesn't make it any less true.

← PreviousPage 2 of 28Next →