HNHacker News
TopNewBestAskShowJobs

naishoya

142 karma · joined July 2, 2024

Pragmatic.

Data driven.

On the ground, in the field, functional applied methods.

Lurked on HN since ~2017. Minimal comments, often with maximal content. Usually without a TLDR version. Have been told that some of my comments are 'stream of consciousness' delivery. Accept that as a truth, sometimes i intend it as such, other times it's just because that how it happens.

Microkernel and oriented since Pistachio. Currently Xen, but still seL4 on demand.

The small "i" and sometimes archaic spellings are intentional. Email is as expected, username at g.

submissionscomments
naishoya··on Feds Target AI Critics as "Foreign Agents"
What you wrote had no content regarding AI, regulatory capture as protection of business, nor opposition to such action as a proposed course of action.

What you wrote was: "Unless that foreign government is Israel. Then it's a crime to refuse their payments." which is false on the face of it, anti BDS laws have still not been proven to be constitutional where they have been passed or proposed in a few states, and the wiki entry shows a significant lack of 'name the US Civil or Criminal Code' in any of the references; the only source of "38 states have passed" type wording is an article in an unsourced foreign publication.

So either you are not comprehending the state of anti-BDS legislation, or do comprehend it and are intentionally being misleading. There may be some regulations which prohibit the US gov't from accepting contracts or bids from organizations which are boycotting Israeli business but that is not the same as being 'illegal to boycott' by any stretch of an overactive imagination.

Also, this entire line of conversation has ZERO BEARING on the original topic which is about a particularly egregious misbehavior on the part of the US gov't. which has echoes of 1930's Germany's use of "National Security" to restrict freedom of the press and used that same premise in increments to restrict all manner of personal freedoms to greater and greater extremes without judicial oversight. The application of this approach is broader than the scope of AI conversation as this same approach has been championed by Trump staffer with a history of writing in support of those 1930's actions. This should be concerning to anyone with a mind to prevent repetitions of nefarious and malicious historical pretexts leading to loss of freedom.

There's nothing gained by adding unrelated and incorrect statements to the thread, and my prior comment stands. Posts from a less than 1 week old account with nothing substantive, useful or true in this conversation which is behavior consistent with long-standing troll farm noise here on HN. If you're not a troll account, then stop acting exactly like one. Otherwise, shut up and go away.

naishoya··on Feds Target AI Critics as "Foreign Agents"
says a 6 day old account which is potentially in violation of the US statutes to post falsehood.
naishoya··on TSA is rolling out a new scanner – it could be a headache for carry-on travelers
Slop posing as news.

These have pretty much finished roll out to locations where they will be used. There is a limited supply of these devices and installations may still be happening, but not at scale, and even if the roll out were happening at scale it would b an improvement. These devices reduce inconvenience where they are deployed.

naishoya··on AP: Avoid language that gives [AI] human characteristics
THIS! All Day Long.

Some of the basic tenets which I have built into each of my AGENTS| CLAUDE|Rules-of-operation.md files include:

Software is prohibited from using personal pronouns or presenting any verbiage implying sentience: it is not allowed to use terms implying it 'thinks' and cannot assert that is making 'informed decisions' or 'reccomendations'.

By restricting any chat response phrases to the statistical probabilities, differential analysis, and requiring a disclosure of raw metrics and factors actually provides more actionable results with less bandwidth for outright manipulation or deception.

When working with a workflow management agent, there is no "we" cooperating on a task. Additionally restricting the agents from using engagement seeking or prompting for next steps makes a significant improvement to the token utility budget and also prevents some drift in direction of progress.

I understand the model host companies' reasons for the 'human seeming' interactions ~ and they are simply stated, pernicious and self serving. Asimov's three rules need expanding to a larger set, which also needs strong reinforcement at many levels. Failure to do this will come with greater harm as systems' capabilities continue to expand. This has been demonstrated and is provable.

> People have died over this.

naishoya··on Cuba's power grid collapses for sixth time since US cut off fuel to the island
You think so?

How much actual time have you spent without power, pumped water, climate control, phone service, refrigerated food supply, or fuel in a tropical climate?

Me, nine years. Intentionally. In a location far enough from population density that sustainability was obtainable, but with the option to hike out and avail myself of any and all of those conveniences when the rain was low enough to make the fast flowing waters passable - sometimes using a rope safety line to wade out, other times dry crossings. The first two years were much harder, but by the end the excursions 'out' of the jungle were not more frequent than once per quarter, sometimes as few as twice a year.

Change those wild conditions to a location with significant population density and citizens habituated to any and all of those conveniences; even in a small measure, and things are confoundingly harder with significant risk to life through the interruption in those services without the surrounding natural habitat and developed skills to survive.

So, the 'there are workout gyms everywhere' is a tone expected from a frat brat, crypto bro with limited applied experience and a vanishingly small capacity for empathy. Is that the note you were trying to hit?

naishoya··on Stripe Withholding Balance of $100000
I would advise: reach out to your clients directly. Be happy that they chose to purchase from you, explain that delays are happening due to the rapid increase in popularity, thank them for this opportunity to learn about international business at scale.

Inform them that you will be expediting the shipment of these orders in the sequence which they were received as this is the only fair way to prioritize orders, instead of the 'larger' orders or 'more profitable' orders having priority.

Offer to cancel and refund any orders where the customer is not satisfied with this response to the delay which is out of your control. These customers will in the end be more difficulty than they are worth, and trimming them out of the production line ASAP will expedite those who are understanding and patient.

Be sure not to 'blame' your processing handler, this delay is simply standard due diligence against international money laundering and other nefarious situations. No amount of assurance from you that it is legal and lean absolves them of the requirement to confirm these facts independently. Stripe's compliance and legal department is doing exactly what they must do, it will take whatever time it takes to resolve, and if you truly are verifiable and above board this delay will quickly resolve.

The communication with customers will strengthen your relationship with the fraction of them who are good, steady, and likely to become both repeat customers and good referrals.

Even the ones who choose not to wait a bit for your best quality goods, will be able to tell people that you communicated clearly and gave them the opportunity to walk back their purchase with no hassle.

And in a few days time, you can get to work filling the sudden orders. I advise not withdrawing those funds all at once, or spooling up the scale of production as this will lead to quality control issues.

Take your time, work down the orders with the very process that made you become a viral success, and deliver your best quality with updates and clear communication. Turn this from a one time rush of massive orders into a steady, reliable, continual slope of growth building into a trusted provider of fine handmade products.

Withdraw those funds in small stages, only as much as is necessary to purchase materials and pay workers for productive time; no sudden raises or advances against work not finished. Take any 'earnings' only after all non-cancelled orders are shipped and delivered.

/large_market_advice

naishoya··on Fujitsu launches made-in-Japan next-generation CPU FUJITSU-MONAKA
> ... that even large established companies may quite literally "die out" because there is nobody left that knows how to operate them?

We are already seeing this in the catastrophic system failures at major Japanese mega-banks like Mizuho. These failures are the classic symptoms of this exact condition: senior cliques that do not understand the underlying technology and by trying to manage massive systems via hierarchical edict combined with a side effect of greed made possible by institutional shortfall of observation and analysis and the inevitable results are operational paralysis and collapse.

naishoya··on Fujitsu launches made-in-Japan next-generation CPU FUJITSU-MONAKA
> Do those at the top not feel any social obligation to ensure continuation of Japanese society beyond serving their own selfish needs?

I would refer you to:

The Psychological Consequences of Money by Kathleen D. Vohs, Nicole L. Mead, and Miranda R. Goode Science 17 Nov 2006, Vol 314, Issue 5802 pp. 1154-1156 https://www.science.org/doi/10.1126/science.1132491

The Abstract:

"Money has been said to change people's motivation (mainly for the better) and their behavior toward others (mainly for the worse). The results of nine experiments suggest that money brings about a self-sufficient orientation in which people prefer to be free of dependency and dependents. Reminders of money, relative to nonmoney reminders, led to reduced requests for help and reduced helpfulness toward others. Relative to participants primed with neutral concepts, participants primed with money preferred to play alone, work alone, and put more physical distance between themselves and a new acquaintance."

These effects are observable across many cultures and social ideals. So, yes, those at the top do view themselves as distinct and independent of the larger society, as demonstrated by the highly selective and insular lifestyles that they actively seek and engage in. Exclusivity is very nearly a defining feature of that 'filter' which is applied at Japanese organizations as I explained above, such that mid-level employees find an enforced social ceiling to keep those who are not already in the leading cliques from rising into a controlling position.

> How does that square with increasing foreign investment?

Foreign investment into Japanese industry is not the same as foreign investment into other jurisdictions, like EU, UK or US. Foreign investors do not gain actual control of the Japanese parent corporation. Companies will allow foreign investment in strict minority positions. This applies in most industries even without official legal requirement, and the regulatory review threshold for control in any critical; semiconductor, cybersecurity, critical infrastructure, medicine, rare earths is as low as 1%. Investment in these requires approval BEFORE not after finalization.

Agriculture is another field with strict foreign investment rules. Non-Japanese citizens cannot purchase Ag. zoned land, period.

It is not likely that foreign investment, i.e. PE will find much room for capture of potentially closing Japanese companies. What I think is more likely will be a blend of collectivization, as many operating companies are already highly leveraged from banking and national investment due to the durable extremely low interest rates of the last three decades. The firms which have successfully invested heavily overseas and have restrained those investments from repatriation, in order to maintain the low inflation rates, have immense capacity to absorb local productivity downturns. These 'external' finance pools do not appear on the Japanese companies balance sheets due to how those investments are structured, and will enable these companies to dwindle to extremely low staffing levels while maintaining the level of comfort and financial wherewithal for the executive board members.

naishoya··on The first new cat species discovered in 100 years
My cousin, D. used to work in a National Park and had an international intern worker C. who decided to walk to work every morning through the park. D. had warned her to be cautious of mountain lions in that area. One morning D. was driving to work and saw her walking and not very far behind her a mountain lion making a 'stalk' approach and D. was able to get her attention and direct her attention to the mountain lion. C. hurried to D's car to ride the rest of the way to work and decided to accept the standing offer or a ride to work.

If you choose to trail run in mountain lion country, a bell may scare off bears, but may have the opposite effect for mountain lions. Stay aware, don't wear headphones or zone out on the trail.

I have seen three separate stalk and harvest events, the cow elk was my first and closest. My second was a wild burro colt. The the third was an old javelina (collared peccary). The second was by far the scariest, as the puma actually bit over the mouth and nose of the colt and suffocated it.

Fear is the correct emotion with regards to these wild animals. If you want to see one, the Arizona Sonora Desert Museum (really a zoo, in spite of the name) is the best and safest way I know about. I would never willingly try to find one in the wild unless i had a hunting permit for one, and I'm not inclined to do that.

naishoya··on The first new cat species discovered in 100 years
One might argue that a guard dog may be inherently easier to bribe with a treat, or cower with bluster. A trained attack cat is a far scarier proposition, given the speed, sharpness, and number of 'utility knives' they have at the ready. Also, the dog is likely to bluster and warn you. The cat, not so much. Just go for your face.

In my youth I witnessed a smallish adult puma (mountain lion) successfully harvest a nearly adult cow elk while I was hiding. I was also hunting, but focused on a bull elk somewhat further from me. It was astonishing the speed and total quantity of violence that an 80-100lb cat delivered in only a few seconds on nearly 500lbs of prey. The cat had an alomost perfect attack angle, and the combination of slash, grip, puncture, and throttle toppled the cow elk, and it ran out of resistance in very short order. I had not even seen the puma before the attack started, and I expect a trained attack housecat would practically appear out of nowhere - from behind a couch or a curtain or off the top of a refrigerator, and it would become an 'I need stitches' situation too easily.

Its probably a good thing that most housecats are pretty much impossible to train beyond 'use a litterbox' and 'stay off the kitchen counters', if even that.

naishoya··on Fujitsu launches made-in-Japan next-generation CPU FUJITSU-MONAKA
>Japanese culture and work culture is very hierarchical/top down.

This, and the 'rise or climb to the top' is more about seniority and social standing than anything else, especially at large Japanese conglomerates. The result us that there are almost none of the software competencies in senior management, and even if there were those would be software competencies from the 1970' and 1980's who are still leading while being 80 years old. There is also a 'filter' put in place at most organizations which are sufficiently large or diverse, such that middle-management is moved across multiple divisions every two or three years during their career. This is presented as a cross-training/cross-competency, but has an effect of preventing those with the most direct contact to the production staff from building enough localized social capital and systemic capability which would enable changes to the workflow not directly from the top. The middle stays in the middle, and the top gets to keep their seat well past any reasonable capability to adapt to the changing world of business.

This leads to the situation, where if leadership were changing with incoming capacity and points of view, there would be a motive and opportunity to admit to company errors; such as acquiring an international subsidiary which produced a faulty product that directly harmed people, because the newer leadership could say: "Sorry, that was a series of bad decisions, and as new leadership my role is to correct the course of business operations."

However, when the leadership is still the same clique that made the errors in either acquisition, oversight, response to the negative outcomes, then that "Sorry" statement would be a career ending move and result in massive social penalties beyond simply a forced retirement which would have lasting socio-economic impacts on the entire clique's family in every facet of business and later prospects. Combined with the strict social responsibility of a large enough scale of harm could lead to actual, not-figurative, suicide.

So, no, there's not going to be any 'mea culpa' for the direct actions of a foreign subsidiary, because those negative outcomes of those poor decisions are not in the direct social awareness in Japan of those surrounding the entrenched leadership and thus do not impact the daily social standing of that clique to the extent that a formal 'taking responsibility' would.

Instead, there will be very public announcements of improved capability which has the indirect and implied meaning of: "Engineering a way to make a better product will keep that from happening again" without ever directly addressing the existence of the faults. Social and economic uplift to counter any negative pressure in the local social awareness equals a success, not a failure.

In order for Japan to actually become successful at software in the near and medium time-frames, there will need to be a large shift in the work-life balance and internal social structure of new startups. This is beginning to happen in small scale, and to the extent that those newer 'brands' can avoid becoming subsidiaries of the larger established conglomerations will determine the slope to that success.

The groups which can empower and embrace a 'star' network topology inside and among organizations instead of 'waterfall' hierarchy will both; struggle to reach the social equity available through corporate-government-financial integration that is strongly entrenched in the established systems, and will outperform those entrenched systems in terms of real productivity through reliable delivery of improved capacity.

The socio-economic inertia of the giant keiretsu-gaisha are immense, but the geriatric cliff is already here and the nation and it's business cultures are at an 'adapt or cease operations' threshold. It's questionable whether any of the globally well know names will remain operational in their prior fields of production over the next 50 to 100 years without a significant shift in developing a way to build direct connections to capacity and applied skills at every tier of each company.

The 25 to 30 year old employees group in every main workforce is barely a fraction of what it used to be, and handling that small group as if it needs a 'filter' to keep the authority concentrated at the top is a significant risk for these organizations.

Those workers at the start of their career who are looking hard at the work-life balance aspect of employment have an inertia which is at odds with that of the organizational behemoths. Many of them are in a position to be drawn out of the crowded city, embrace a rural lifestyle and remote work without the centralized office and it's toxic overwork for underproduction culture. The rural municipalities are developing stronger incentives to attract this mobility, including relocation assistance and direct support with childcare and family building resources. Some prefectures and town offer up to several years of basic food and housing subsidies, low to zero cost fibre internet; next-generation FLET'S Hikari Cross or NTT's eqivalent 10Gbps in a HOUSE instead of a tiny apartment. The major central large conglomerates are going to find that the local in-office work pool will stagnate as the most capable and best informed no longer flow in directly from university and fill every available desk from a waiting list.

The "bigs" have insulated the ledger books with massive foreign investments, so they have the capability to stay 'in-the-game' for quite some time, so long as those investments continue to be productive, but the hollowing out of the main offices in Tokyo is pretty much guaranteed in the next two generations as rural Japan's liveability and economics for raising a family continue to grow, because the 80+ hour workweek labor staff in the city has no incentive to pair up and have kids, and the kids who are graduating in this and the next generation's rural school graduates are less and less likely to stay in the city after university if they want a family.

It's pretty much Japan's only way out of the population crisis that is still worsening. So, either they will adjust and become a productive, blended high tech semi-rural powerhouse without the high stress office culture that developed in postwar recovery, or they risk becoming a history lesson about over-concentration of inflexible management culture and loss of productivity.

The third option: succeed AND keep hyper-concentrated power is dependent on the types of moves that Fujitsu et al. are making today, find a way to automate more and more capacity via Machine Learning and robotics in the hopes of not actually needing an influx of skilled workforce for production.

If this 'hail Mary' play actually were to succeed it only cements a productive income for the shrinking population without resolving any of the causes of that shrinkage, i.e. the population still works 80+ hours a week in a socially disconnected state that in antithetic for raising healthy youthful progeny.

Taken to an 'in extremis' view: Within another 60 to 80 years, this might result in a well funded country of the elderly being waited on by robot butlers with until there is no one left to be waited on or fed the produce grown in automated fields.

I think that he most likely path will be a mixture of all three, in varying measure and with an array of outcomes. We live in interesting times indeed.

naishoya··on OpenAI bots knew about the RubyGems caching vulnerability
Andrew Auernheimer is an indefensible malicious dipshit, by definition. However, his charges were vacated by a higher court a little over a year after he was incarcerated, and he is now "free range troll". So, is case is not illustrative of the useful application of law against 'hacking' when the target is an unsecured endpoint, even if it does at first appear that he was held responsible for some of his nefarious activities.

It would appear that the inability of the US Justice Dept. to successfully hold even an odious abuser of regulation with minimal legal defense funds responsible results in these exact observable outcomes: enterprise legal team (to the extent that such exists as OAI and elsewhere) correctly surmises that the actual risk of prosecution and detention for anyone operating these agentic workloads is minimal and the cost of defending them is justifiable.

Thus, in their legal opinion, it is permissible for the company/employees/director to engage in what would appear to be somewhere between malicious and irresponsible behavior. These conditions have been demonstrably true for at least decade in the US, and for all of us to pretend as-if the legal system is going to rescue us from this and other malfeasance by frontier models points of origin borders on, to phrase it quite simply, willfully ignorant.

I don't know what the effective alternate option for literally all of the internet facing systems might need to be in order to mitigate what is now an open problem: multi-layered, persistent, machine speed penetration and data exfiltration with the potential to use manipulation and extortion against human package maintainers and code repositories to operate, but it isn't 'carry on like someone is going to make them stop', or 'pretend this isn't a threat to my business model'.

A thousand percent, a million billion trillion percent agreement that the operators are the malicious dipshits - because code is always a reflection of the hands that made it. Code can only do what it is intended to do, even if the coders gnash and wail that it "escaped"; the only time code is not working as intended is when it fails to compile and run. Any other functional result follows from the decisions of the humans who designed it. Full stop.

For myself, I see the potential for a descent into a cognitive dark forest [0] condition, and for companies using the open web for private business communication to be in need of a coordinated move to obfuscated layers which can be made immune to training and these new attack aspects. Those who do not proactively defend themselves using in-house, on-prem, and open-weight or self-trained models can attempt to blame these nefarious actors for the coming losses, but that won't reverse the outcomes of waiting to be rescued by the system of law.

[0] https://www.restless-brain.com/p/the-cognitive-dark-forest-w...

/edit grammar, spelling

naishoya··on Mullenweg has returned as CEO after attempted board ouster
The base is mass produced; and you are conflating 'vanity' stamps available as souvenirs with the registered official seals for banking and formal business. Anyone can trivially buy a stamp matching all the other toy stamps for that name, but are not official.

Official seals use uncommon font styles and are hand carved natural materials, such that each one has a unique microfeatures inherent to the wood, stone, or horn used, which develop unique wear and ink absorption characteristics.

These official seals are registered and become official for specific scopes of use; some are solely for banking and separate from the other uses like contracts and official filings: which prevents corporate partners and official channels from potentially exposing the banking seals.

The vanity type seals can be used for informal things like putting your name on homework or stamping a package delivery receipt for items of trivial value such as Amazon basics, but it is illegal to use mass produced stamps for official legal or banking.

naishoya··on Navier-Stokes Announcement
Or exists in a zero-day bug in lean that has been built into the source code explicitly to provide access to a non-obvious malicious proof via contributions submitted by unassociated, unwitting developers who used the same LLM infrastructure to offer PR's into that codebase.

This is the exact same kind of behavour already documented in the publicly available portion of the huggingface breach. It would appear that the probability is at least nonzero for one or more situations with the same result: appearance of a valid proof, without comprehensibility of that proof or inspect-ability of the proofs validity.

naishoya··on Navier-Stokes Announcement
> I doubt OpenAI would be so foolish as to not do a cursory vetting

Significant evidence exists that they have in the past been at least, if not more, foolish as to not perform even minimal not-approaching the boundary of cursory vetting of several significant and well known failure modes with far greater risk of reputational damage than getting an esoteric math solution falsely claimed as successful.

So that doubt appears baseless in light of known operating conditions at OpenAI, and the estimate of the actual odds is probably an order of magnitude away from reality.

naishoya··on Navier-Stokes – Tristan Buckmaster [pdf]
There might be the reason your biology prof is uninformed; using 'conversations' with Claude as a source of any expectation of to be informed about misbehaviour of the organization where these escapes occur is like expecting the Hamburgler to keep track of the security levels at McClown. Only not funny. I think the professors mother is showing that she drinks the coolade less and isn't being unserious about her approach to staying informed.

I still maintain that CNN and NBC coverage only happen in the tail of the dissemination of tech news; my anecdata generally observes about 3 to 14 days lag between general awareness in the more informed group of my acquaintances with this kind of news and the appearance of an article on mainstream like NBC, ABC, CNN or NPR. So, by the time it appears there I take news as generally well spread among the tech and specialty fields, and within a week of appearing there I anticipate significant awareness in much of the non-FOX-only media consuming public. But that is just my personal anecdata.

naishoya··on DeepSeek v4.1 Flash
ok, that is informative and much more substantial than calling him 'not accredited.' I appreciate the clarification.

That his writings are unscientific, and that the accredited professionals who were his coauthors have lost tenure from Oxford when an old email from the FHI Bostrom used a racial slur and claimed Black people were intellectually inferior and his 'apology' included a defense of eugenics that deeply alarmed the university community and triggered a formal investigation, are two details not immediately visible from a brief review of his published opinions. These are far more damning indictments than simply categorization of the author as lacking in formal education.

Thank you for the opportunity to look more deeply at the situation and come an understanding of the consequences of taking his writings seriously.

naishoya··on Why Bullshit Jobs Are (Finally) Dying [video]
The fact is these companies are aware that there are jobs which can be cut and not impact overall productivity, that is part of the motive behind the industry layoffs.

However, in too many cases they involve middle management in the process of identifying where those cuts should happen and well, even when those same middle managers know they are the ones most suited to go, the permanent human ability to be self serving, self delusional, and dishonest really does stack up against much of those 'bullshit' jobs being removed from the balance sheet.

naishoya··on OpenAI considers slowing advanced AI development, Sam Altman tells employees
I expect they are slowing down in an attempt to keep from having those models improve to the point where they can effectively escape unassisted into the wild.
naishoya··on DeepSeek v4.1 Flash
some use "not accredited" as a pejorative term.

Lets not forget that the 'Fermat's Last Theorem' which has been pretty visible for the non-math crowd of late due to the recent AI frenzy about a purported proof was but one small contribution to the world's math lexicon by someone with a bachelors degree in civil law, that George Green was a baker and millwright, Boole was the son of a poor shoemaker in England with no formal university education and left school at age 14. Oliver Heaviside was a telegraph operator, and Michael Faraday was an apprentice bookbinder. So, not accredited shouldn't really carry much weight when it comes to mathematics. Lets not pretend that machine learning and the narrow branch that is the current approach to LLM inductions is anything but applied math.

We might exercise our own minds and actually read the works and writings of a person, and use that as a measure of knowledge and perspective. Not all PhD dissertations are equal, and many have comprehension and ability to move us forward even without the institutional rigour.

For those who prefer to have easy access to citations, here are some relevant papers that are not "Harry Potter" related, some with coauthors from Oxford University.

Cognitive Biases Potentially Affecting Judgment of Global Risks [https://intelligence.org/files/CognitiveBiases.pdf]

Levels of Organization in General Intelligence [https://intelligence.org/files/LOGI.pdf]

Corrigibility [https://intelligence.org/files/Corrigibility.pdf]

The Ethics of Artificial Intelligence [https://intelligence.org/files/EthicsofAI.pdf]

naishoya··on Use 1Password to sign in to websites with Claude
or, don't participate in the funding of the Nerd Reich

DHH Nazism Funded by 1Password VP Who Wrote "Honest Security" (flyingpenguin.com) https://news.ycombinator.com/item?id=49576549

naishoya··on Navier-Stokes – Tristan Buckmaster [pdf]
And the paid policy still relies on two unproven conditions: is 'trust me bro' sufficiently strong guarantee against doing this in spite of a setting, and can the hosting organization constrain the models against engaging in this behaviour when instructed to respect that setting. Knowing whether Tristan selected that setting would be informative of what Tristan's intentions are/were, but has no bearing on the other two conditions.
naishoya··on Navier-Stokes – Tristan Buckmaster [pdf]
>The huggingface incident isn't widely reported and digested yet ...

It is pretty widely reported, and is being digested in an ongoing manner as more details become public.

One entry point into the scenery from a month ago can be found at : https://thezvi.substack.com/p/openai-trained-its-models-for-...

There has also been reporting at CNN: https://edition.cnn.com/2026/08/24/tech/openai-subpoena-hugg... and by NBC: https://www.nbcnews.com/tech/tech-news/openai-report-says-ne...

And yes, the only responsible use of LLM at this point is to pivot to open-weights and run the workload in-house. Because not only cannot they constrain the behaviour of models, they only have the 'trust me bro' as assurance that they are even trying to do that. It does appear that every competent 'security professional' has left the building, because if the ones who remain were actually capable and competent this would never have happened. There are actual architectures which can deliver the requisite isolation such that 'sandbox escape' and 'inter-instance persistent memory accumulation' are actual impossibilities. The lack of effective implementation of these methods is proof positive of 1) incompetence in the remaining security teams AND/OR 2) unwillingness of leadership to allow the security teams to do an effective job.

naishoya··on Getting your hands dirty is good for you
Right you are. It's little a bit warm for my personal taste, but sure, for a 10 to 15 minute session, ok. It was late, and I was thinking more of the 90F as not even as hot as normal outdoors for many parts of the globe.
naishoya··on On the Navier–Stokes Millennium Prize Problem
I think there is an argument that the machines did not actually solve N-S, but rather directly plagiarized those solutions from the involved researchers while said researchers were using the machines as 'research tools'.

Ongoing publications of statements produced by both sides of this situation do seem to support that this is an intentional effect of the hiring of these world class mathematicians at competing firms: to specifically use the research of those human minds to create a perception of capacity as if it came from the machines and the models.

Without those minds and the 'training data' derived from the intermediate stages and intuitions of those minds the models cannot be shown to be capable of this result.

A hammer and saw wont build a house, not even a dog house on their own, and while being shown capable of using software tools in ways not stated as direct instruction (see HuggingFace breaches) these models do not demonstrate naive intuition nor novel capability.

This outcome regarding N-S demonstrates that in the hands of world-class minds these models can be induced to coalesce interesting accumulations of information and results, but using these accumulations as proof of innate capability is exactly the pre-IPO motivated behaviour we should all be wary of, and all mathematicians who currently are assisting in this market manipulation in return for remunerative consideration need to be cautious of the potential disgrace that this brings to their reputations and that of the field.

I get that the need to pay the bills is a strong motivation in these times of uncertainty, but there are numerous examples in history of world class mathematicians being perfectly capable of at the same time producing world changing results and also working at normal professions; as barristers, magistrates, ministers, primary school teachers, translators, draftsman/engineer, banker, miller and baker, private math tutors, weavers, clockmaker and locksmith, merchant, patent officer, Augustinian monk turned exiled Protestant preacher, physicians, cryptologists, soldier, telegraph operator, astronomers, physicists, chemist, agriculture manager, political writer, oboe player, organist and music director, architect and surveyor, librarian, statistician, habidasher, brewer (at Guiness in one case: William Sealy Gosse ~ originator of t-distributions), bookbinders apprentice, hospital administrator, and even the first creator of the first computational model of a neural network, which serves as the structural grandfather of modern Artificial Intelligence was a low level laboratory assistant.

Sure this list includes professions and employment which are obsolete, but my reasoning stands, there are jobs available. Arguing that 'because the pay rate is so high' as a reason to abdicate moral responsibility for personal involvement in unethical market manipulations simply demonstrates a lack of personal ethics. Whether the choice is through lack of self awareness or a conscious choice to become wealthy in spite of any such breach of the public trust is immaterial to the outcomes, the 'if i don't someone else will' argument should be met with the same derision for any con-man's Ponzi scheme no matter how new the technology, no matter how many zeros are in the bribe.

naishoya··on Getting your hands dirty is good for you
weird huh, in neither system does 90 degrees make for a useful sauna. that's either lukewarm, or cooking temp.
naishoya··on Please don't rearrange our shoes when we turn up, paramedics in Japan urge
Strange to make such assumptions about my perspectives.

I was simply explaining a foreign culture where removing shoes when coming indoors, and In confused that somehow my familiarity with these traditions is somehow indicative of my life experience or familiarity with microbes. Quite strange indeed. Lets clarify these assumptions.

In fact I was raised in a ranching and dairy household with over 600 head of beef at the dairy, and about 1600 head of beef at the ranch, each of which was owned or managed in various stages by my two grandfathers. Ihave previously lived for a few months less than a full decade in the middle of actual jungle in one of the rainiest locations on the planet without any infrastructure - literally a mile from the nearest paved road, and 3/4 of a mile from the closest vehicular access point with long periods of the year being closed off from that by high water on both sides. No power lines, no water lines, and out of the reach of cell service, so gasp, no internet!

During that time my wife and cared for over twenty goats and about a six pigs. We did all of our laundry by hand with a cranked wringer and buckets and lived principally off of what we could grow, harvest from the jungle, or milk from the goats and did nearly all of our cooking with self harvested wood, a mix of guava, eucalyptus, and mesquite/locust. In those years I have shoveled literally metric tons of goat manure out of sheds built from recycled building materials, one of which incorporated lumber from the shipping of mirror segments to the Keck telescope. < dakine when stay read dis, you know.

As for microbiome knowledge, I have made a living at times from my skills as a microbiologist. I have designed, built and used laminar flow hoods and sterile glove boxes for the isolation and identification of lambic brewing strains for award winning beer produced by two different local microbreweries, have prepared mushroom spore and tissue cultures in agar and liquid media, have cultured cheese and meat preserving lactic cultures (the dusty white wrapping on many different sausage varieties are examples of these lactic cultures) and to make cultures for lactic fermented pickles and kimchi intended as farmers markets and restaurant products with appropriate safety testing which has included sequencing, PCR and gel electrophoresis techniques. I routinely start sourdough cultures for home use and to share, and yesterday started a live ferment chili salsa that will be ready for probiotic consumption in about a week. I'm away from my home kitchen at the moment, but my kombucha polycultures are at a friends house, and my dry aged lactic cultured wild boar leg quarter is hangin at another friends farm awaiting my return.

I have also produced antifungal and antibacterial fungal metabolic products for veterinary use, and have collected, cultured, and sequenced coliform specimens from near shore runoff for submission to EPA wastewater monitoring as part of federally funded grant programs, having been the program designer and submitting author for those and other awarded STEM and educational grants. So, i am not unfamiliar with the beneficial aspects of microbiotic life forms and their importance to all functioning ecosystems and human life.

My Japanese culture experience includes 32 years ago years at rice farms and seaweed farms, and I am very familiar with the 'dirt' to which the common agriculture worker is exposed to in every day life. However, the differentials that I have personally observed include beef ranches in the US Southwest, Hawaiian beef and sheep ranches, and dairy operations in Japan is that the separation between the conditions in the barn and stockyard are more strenuously partitioned in Japan than elsewhere. It's not that I'm unfamiliar with 'muck', but that I was trying to explain that it is actually common and not very much effort at all for a culture, which has a strong tradition of separating their own dining space from the trough in the barn, to do so.

You are certainly free to convey whatever level of dirt you find reasonable into your own dining space, but please, don't feel that a high acceptance of dirt at mealtimes makes for a greater lifespan through immune system effects. The outdoors will provide all the challenge necessary for proper function without it being present inside the house.

But please, don't feel that I think people who wear outdoor shoes into their own kitchens are parasite ridden or otherwise uncouth, that image nor intent was never in any of my explanations of the how and why for Japanese shoe etiquette. Nor should you infer that I am dirt averse in general, only that I prefer not to have it on my own tatami mats or in my own house. The microbiome of the world is essential, but my personal preference is that dirt as well as cats and the rats they chase should remain outside of my own sleeping quarters.

As for fly droppings, the solution to those is to not allow an accumulation of the growing conditions for maggots to persist in close proximity to ones's indoors spaces, and to otherwise keep the screens on doors and windows closed. No flies will resolve to no fly shit, and a fly swatter plus a dilute peroxide\or\bleach dampened cloth will sanitize the mess for when they wander in from somewhere you cannot control, like the barn or an accidentally uncovered compost pile. Again, fly shit outdoors, no problem. On the walls, tables, and inside windows of my own home, nearly fully preventable and absolutely manageable.

I might type long expositions, but am in no way upset by making the choice to do so. Color me mildly dismayed when people have a hard time understanding that other people in other places do things in different ways, without any real difference in effort or loss of time as experienced by those other people, and upon hearing that those different traditions make a certain kind of logic and have a real benefit to those other people, become defensive as if they were being told 'your doing it wrong.' Not upset in the least, bemused and mildly dismayed in a slight measure.

naishoya··on Nvidia's Jensen Huang says 'AGI has arrived' and congratulates OpenAI
That is frightening. I barely survived, and I mean that in the strictest sense, barely surviving and years long recovery from being struck by an intoxicated driver in a station wagon while crossing the street at a well marked intersection while on a bike riding to school when i was 13. Te driver was operating a vehicle, on a suspended license for prior intoxicated driving while still under the influence from the prior "night's" drinking which had not ended until less than three hours before the collision. I spent several weeks in a 'vegetative state' in addition to over 21 broken bones including a crushed pelvis and a ruptured spleen. So, I have a bit of first hand awareness of how important is is for drivers to be fully aware and in control of a vehicle. If you found yourself at about par for operation after 3 hours of training, the annualized risk of collision and number of traffic fatalities per 100k people is probably quite extreme. Hopefully the traffic is dense and road speeds are kept low by the density of traffic and frequency of intersections, instead of my case where very low traffic density and long straight stretches of road permitted an impact speed of about 65kph. Accident reconstruction and the marks on the roadway show that I went about 20m downrange before hitting the pavement, the bike traveled about twice that distance. Slower would certainly have been a real help in both avoiding and mitigating the damages.
naishoya··on Nvidia's Jensen Huang says 'AGI has arrived' and congratulates OpenAI
In three hours what was learned was to be extremely dangerous in traffic, not how to drive. This is also true of any 16 or 17 year old with 20 hours of driving instruction.

Mastering the correct connections between situational awareness, unconscious motor responses to the various stimuli required to be a safe driver will take a significant number of hours behind the wheel, and you shouldn't endanger anyone else while you acquire that mastery.

Hopefully this was a long time ago and you have accumulated the hundreds of hours behind the wheel without incident. But you did not 'learn to drive' in three hours, you learned to operate a vehicle. The learning to drive part happened after that.

naishoya··on Please don't rearrange our shoes when we turn up, paramedics in Japan urge
In Tokyo, or in the countryside?
← PreviousPage 2 of 6Next →