HNHacker News
TopNewBestAskShowJobs

gojomo

30,216 karma · joined February 24, 2007

Gordon Mohr, maker of software.

Twitter: http://twitter.com/gojomo

Project: http://thunkpedia.org

Idea blog: http://memesteading.com

Older blog: http://gojomo.blogspot.com

Homepage: http://xavvy.com (username @ here for email contact)

My HN peeve is formulaic downbeat comments, like: "How is this news, I already knew this!" "…Betteridge's Law…" "I stopped reading at…"

submissionscomments
gojomo··on Company Reminder for Everyone to Talk Nicely About the Giant Plagiarism Machine
There is no de jure legal requirement that the RIAA, Disney, Nintendo, or the government be "pleased to hear" about new technology.

And, while copyright prohibits some sorts of reproduction of copyrighted materials, it doesn't give rightsholders veto power over all downstream uses of legal copies.

gojomo··on ChatGPT could never get a PhD in geography
The giveaway that Marcus is generating slop for consumption by unsophisticated AI-haters is that he doesn't bother to mention what models/options involved.
gojomo··on Show HN: Undetectag, track stolen items with AirTag
After I saw third-party "10 year battery enclosure" offerings for AirTags, was wondering when other workaround customizations like this might appear.

Other impactful variants might be:

* senses whether another 'sibling' AirTag is present, if so, stays off. If not, waits X hours & then turns on.

* has its own motion sensor; only after X minutes of being stationary, it waits Y hours to turn on briefly

* has its own clock & (original-user-known) randomization seed; turns on at pseudorandom intervals the original user can predict

* low-power/low-bandwidth receivers so cheap & tiny now: could wait for national or even global unit-specific 'wake' request - perhaps even with parameters for duration/intervals – before powering-on AirTag portion

gojomo··on Launch HN: Tinfoil (YC X25): Verifiable Privacy for Cloud AI
Is there a frozen client that someone could audit for assurance, then repeatedly use with your TEE-hosted backend?

If instead users must use your web-served client code each time, you could subtly alter that over time or per-user, in ways unlikely to be detected by casual users – who'd then again be required to trust you (Tinfoil), rather than the goal on only having to trust the design & chip-manufacturer.

gojomo··on Emergent Misalignment: Narrow Finetuning Can Produce Broadly Misaligned LLMs
Prior discussion when the paper was 1st reported in February: https://news.ycombinator.com/item?id=43176553
gojomo··on Lossless LLM compression for efficient GPU inference via dynamic-length float
Some might prefer the fidelity of this method's 70% savings over the lossyness of 4-bit quantization's 75%.

And, maybe the methods stack for those willing to trade both costs for the smallest representation.

gojomo··on A new form of verification on Bluesky
In the core team's clients, if the 'verified' account changes its display-name and/or handle, does the blue check stay, disappear, or do some secret third thing?
gojomo··on Show HN: HN Watercooler – listen to HN threads as an audio conversation
Can I upload my own voiceprint so my comments are said in my voice, voice of my choosing?

Can I navigate by voice commands, for example if listening while driving?

gojomo··on George Foreman has died
Who's the 'she' in this story?
gojomo··on Hallucinations in code are the least dangerous form of LLM mistakes
Such "hallucinations" can also be plausible & useful APIs that oughtta exist – de facto feature requests.
gojomo··on The San Francisco stadium that fog killed
Yep - KFOG wisely noticed late May offered better atmospheric conditions for an over-th-bay fireworks show, and thus outdid July 4th's visibility most years.

The Bay-to-Breakers fun run takes advantage of the same span of especially sunny SF Bay weather – so for many years the one-two punch of Bay-to-Breakers one Sunday, then KFOG Kaboom the next Staurday made for a very special SF week in the 2nd half of May.

gojomo··on Cities can cost effectively start their own utilities
Of course it's not the 1:1 expense passthrough of PG&E's multidecade sponsorship of just Newsom's family that alone causes higher rates.

It's the tens of millions in political contributions to not just Newsom at every stage if his career but also his allies & ilk.

Plus tens of millions more in donations to favored "charitable" projects of family & cronies of politicians like Newsom & his allies.

Plus harder-to-tally overpriced jobs & contracts steered to the "right" (regime-aligned) consultants, firms, & unions. (Did you notice how the glowing character reference in the Wapo article, about how immune to influence Newsom is, comes from a former Newsom advisor – & member of his wife's nonprofit board – who currently has PG&E as a paying client of his PR firm?)

All the cash & favors are then topped off with a showy cheerfulness to "accept the responsibility" of extra costs from those state-idiosyncratic "regulations" you mention.

But of course that's the quid pro quo of their monopoly remaining unchallenged, & all costs more-than-passed-on to ratepayers – so the ruling machine including Newsom can keep getting their beaks wet in all PG&E's spending & donation fountains.

To look at only the formal campaign contributions against revenues, & accept as excuse a cover-story for extra corrupt expenditures ("regulations are very different"), & then conclude "no problem" is exactly why milking ratepayers this way, spreading the loot widely, is such a enduringly great racket.

gojomo··on Cities can cost effectively start their own utilities
PG&E's absurdly high rates are how they finance the careers of Governor Newsom & his wife, and "donations" to a giant array of other pet project "charities" for politicians & their favored cronies/causes.

https://www.washingtonpost.com/business/2019/11/11/pge-helpe...

https://www.sacbee.com/article251851903.html

Why would any of California's uniparty establishment want to refund all that money to random utility customers – who might not even be reliable party footsoldiers?

PG&E's lowest overnight rates are 30¢/kwh, surging during peak hours to rates ranging from 39¢/kwh to 72¢/kwh.

In neighboring Nevada, the utility serving its major metros NV Energy has rates of 11¢/kwh for residential users & 7-9¢/kwh for small businesses.

gojomo··on US newspapers are deleting old crime stories, offering subjects a 'clean slate'
Reminds me of a laugh-out-loud closing paragraph in a 2009 NYT article about the clash between Wikipedia & German lawyers over some infamous murderers' right-to-be-forgotten.

https://www.nytimes.com/2009/11/13/us/13wiki.html

Read the whole thing for maximum effect, but for me it beautifully demonstrates the contrast between the USA/Wikipedia/NYT ethos of "the truth is always printable and your speech is by default 'on the record'" and alternate expectations elsewhere.

gojomo··on Apple Photos phones home on iOS 18 and macOS 15
My thoughts exactly: "we've got this crafty image fingerprinting, the CSAM detection use proved too controversial to roll out, but let's get the core flows into something that sounds useful for users, so the code atays alive, improving, & ready for future expansion."

Whether such fingerprinting can reliably be limited to public "landmarks" is an interesting question, dependent on unclear implementation details.

Even if the user-visible search is limited to 'landmarks', does the process pre-create (even if only on-device) fingerprints of many other things as well? If so, it suddenly becomes possible for briefly-active non-persistent malware to instantly find images of interest without the wider access & additional processing it'd otherwise take.

gojomo··on Ozempic increases risk of debilitating eye condition: studies
Such a curiously-worded headline, to refer to Ozempic specifically – quite a recognizable & understood term at this point! – as just some "Danish diabetes medicine".

I almost expected the source to be some group with a resentment towards Denmark – maybe Greenland separatists? – but no, it's just a Danish university that doesn't want to say 'Ozempic' in the headline directly.

gojomo··on Gukesh becomes the youngest chess world champion in history
He's 18.
gojomo··on CDC Confirms H5N1 Bird Flu Infection in California Child: First Child Case in US
Wastewater detection is showing H5 bird flu all over California: https://www.cdc.gov/nwss/rv/wwd-h5.html

That's suggestive there are far more than the 29 reported California cases - with cases generally remaining mild enough to earn no special attention over routine respiratory infections.

gojomo··on FDA proposes ending use of oral phenylephrine as OTC nasal decongestant
In my experience oxymetazoline has similar fast-action - even "10 minutes" seems a bit on the long side.
gojomo··on FDA proposes ending use of oral phenylephrine as OTC nasal decongestant
Is there any basis to think xylo- is better than the similar oxymetazoline available in the US? Both the efficacy and downsides seem similar from discussion so far.
gojomo··on New York Times Tech Guild goes on strike
Can an employee quit on a whim without any reason, taking vital functions away from the productive team on which they served?
gojomo··on The Saga of a Celebrated Scientist – and His Rodent Dystopia
Gwern's deep dive into Calhoun's scholarship suggests taking his "results", & popular/stylized glosses thereof, with some pretty big grains of salt - in line with the fall from academic prominence described in later parts of this article. Gwern's abstract:

> Did John Calhoun’s 1960s Mouse Utopia really show that animal (and human) populations will expand to arbitrary densities, creating socially-driven pathology and collapse? Reasons for doubt.

https://gwern.net/mouse-utopia

gojomo··on Court tells EPA to consider fluoride risk
Not all sunshine would be 'medication', but if a doctor recommended an extra hour in the noonday sun (or equivalent time under appropriate lamps) over a patient's baseline, that extra increment of light would be 'medication'.
gojomo··on Court tells EPA to consider fluoride risk
Yea, all the costs and benefits should be weighed!

I suspect a lot more people die over their lifetime from being -5 IQ than cavities. But, let's be sure all the effects are net-tallied – not just an unquantified hand-wavy "saves lives via dental health" with zero consideration to any stupidity or behavioral issues created.

gojomo··on Court tells EPA to consider fluoride risk
>The safe levels are built around a model of the amount of water a human being can consume, both directly (drinking tap water) and indirectly (cooking, drinks made with tap water, etc.).

The epidemiological studies suggesting detectable harm at 1.5mg/L are from whatever amounts are actually consumed in affected areas, not some "model" of the max any person could consume.

Given the variety of human practices – preferences for packaged drinks, different food-preparation approaches, etc – the same tap levels could easily result in gigantic differences in net consumption among different individuals.

If the standard is half the "safe" level, but some people consume 4x or 10x as much tap water, & perhaps even have a higher baseline of non-tap-water flouride consumption, then some percentage N of the population could be developmentally-harmed, even at levels "safe" for "most". How high of an N should we accept for the modest benefit of fewer cavities?

Your link shows no public benefits other than less tooth decay, doesn't compare flouridation to other interventions, & doesn't weigh costs versus benefits.

If I'm reading its "DMFT Index" ("Decayed, Missing, Filled Teeth [per capita]") values right, it also shows a very modest benefit: a "DMFT" of 0.06 in flouridated areas, and 0.115 in non-flouridated areas. Roughly, instead of one cavity-affected-tooth per 17 kids, there's one cavity-affected-tooth per 9 kids.

Yes, we should try to minimize cavities, but what N% of outlier-consumption overflouridated kids should get a 5-IQ-point drop for that marginal benefit?

What if we shifted the flouridation expense into reminding kids to brush or use a flouridated rinse? What if we just paid a cash bonus to kids who manage to avoid permanent-tooth cavities through whatever interventions work? (The same funds would go to treatment/amelioration for those who wind up with decay anyway.)

gojomo··on Court tells EPA to consider fluoride risk
As "medicating" implies addition, I'd not call such filtering "medicating", but it's definitely still be a "medical intervention", a purposeful change from the default motivated by health concerns.
gojomo··on Court tells EPA to consider fluoride risk
https://www.theguardian.com/society/article/2024/aug/23/fluo...

> The long-awaited report released this week comes from the National Toxicology Program, part of the Department of Health and Human Services (HHS). It summarizes a review of studies, conducted in Canada, China, India, Iran, Pakistan and Mexico, and concludes that drinking water containing more than 1.5 milligrams of fluoride per liter is consistently associated with lower IQs in kids.

>The report did not try to quantify exactly how many IQ points might be lost at different levels of fluoride exposure. But some of the studies reviewed in the report suggested IQ was two to five points lower in children who had had higher exposures.

>Since 2015, federal health officials have recommended a fluoridation level of 0.7 milligrams per liter of water, and for five decades before the recommended upper range was 1.2. The World Health Organization has set a safe limit for fluoride in drinking water of 1.5.

gojomo··on Court tells EPA to consider fluoride risk
Taking assurance from the fact that municipal-flouridation goes to "half or less" some "generally recognized safe level" implies that a low-concentration is enough for safety – "ensure your water's at that level, then have as much as you want!" – but that doesn't seem to be true.

It's total consumption that brings risks. Someone who drinks twice as much water at that presumed "safe" level is at more danger – as are those who might be getting extra flouride elsewhere, such as from higher tea consumption or excessive use of flouridated toothpastes.

Pre-2015, US flouridation standards used to be higher in colder climates, and lower in hotter climates, to account for the added water consumption in hotter climates. As of 2015, they guided everywhere to the originally-only-for-hotter-areas lower level, 0.7 mg/L, which is about half the levels with the decent evidence of harm.

But that could still be putting high-total-consumption people at unnecessary risk, and for a pretty-small benefit (fewer cavities) that can be obtained with non-ingesting tooth surface flouride exposure (as with toothpaste).

Any anti-flouridation advocate who relies on "their groundwater" rather than municipal taps probably does care about its natural concentration levels, and would only have to be concerned with moderate tea drinking if they're also receiving extra surprise flouride via municipal water. (A authentic concern for minimizing total consumption doesn't make any particular neurotic avoidance a good litmus test.)

gojomo··on Court tells EPA to consider fluoride risk
Adding something explicitly for medical benefits, because it was recommended by medical professionals, is "medication", yes.
gojomo··on The Optimus robots at Tesla's Cybercab event were humans in disguise
Why humanly shaped, you ask?

Drop-in compatibility with all human-shaped legacy interfaces, for the most-rapid deprecation of homo sapiens as soon as the ASI can teleoperate humanoid proxies.

← PreviousPage 4 of 34Next →