HNHacker News
TopNewBestAskShowJobs

w10-1

3,773 karma · joined January 31, 2022

submissionscomments
w10-1··on An Alien Mind
> the best argument for AI is that it's an arms race

It's best not to reduce AI momentum to arguments, especially the "best" arguments (meaning I suppose most acceptable?).

The same forces that feed and motivate humans and that drive resource and governance decisions generally also strongly support building AI, particularly insofar as it can deliver strategic advantages in our many competitions over resources and influence. Cyber-defensive use is at best a nice side-effect, but itself might be cast aside for the sake of other advantages.

In this historical moment, due to the need to generate public interest in product, equity, and debt offerings, some of this building happens in the open. But the military-industrial complex prizes secrecy, in part to hide capabilities, but mostly to imply more capabilities than they actually have. Historically, critical innovation will get bottled up in secrecy (which not coincidentally gives them the power to choose who will gain), but frankly that market is much smaller than enterprise and consumer. So we can bet that it's not only "open models" that are targeted to go under wraps, and more broadly we should not believe that the intentions of researchers matter, but whether governments are more interested in the strategic benefits than the economic ones (or view the economic ones as net-negative for their jurisdications).

w10-1··on The Rise and Fall of Agent Civilizations
> the evolution of more efficient communication systems because it would require going through a stage where communication between signalers and receivers is perturbed

"Worse is better"

w10-1··on California lawmakers unanimously pass Linux exemption from age-verification law
> copy, redistribute, and modify the software

I can copy and redistribute macOS binaries, and I can write programs/extensions that modify macOS.

These vague terms show that legislators are incapable of regulating software effectively; but they do create an enduring legal franchise to deal with their confusion.

w10-1··on The Teaser Period: Why the AI Boom Is Hitting a Reset Wall
> large capital investors [...] will be very highly motivated not to let their investment be seized [...] So the labs will not have too much difficulty raising or borrowing enough money to pay the bills

So: sunk costs of large investors mean they'd be willing to pay high interest costs? Thus small investors can rely on this dynamic, shielded by big ones?

Remember, the large investors got large in a near-zero interest rate environment (where exact timing doesn't matter as much), but those days are not coming back. Soon margins will matter most, and while NVDA and Apple have the internal discipline for that, OpenAI et al do not.

w10-1··on Apple introduces M6 and M5 Ultra
Leasing now an option, only $50/month (cheaper than inference subscription?), so even cash-poor can go the amortized-investment route.

I've often felt there is tremendous value locked up in underutilized old computers. It would be interesting to see Apple in 3 years offering compute as a service using lease returns (or more likely, partnering with someone else to operate it (perhaps exclusively in secondary markets like China or India, to address political demands for local siting or jobs). Apple is in the best position to work around or even gap-fix older software/hardware limitations in a controlled environment, and now they can do so without cannibalizing new hardware sales.

w10-1··on How credit card rewards became a $9.2B wealth transfer
The study methods are closer to advocacy than science or policy.

Sure, take any slice of a vast number, and you get a big number.

It's not a "wealth transfer" when everyone gets what they bargained for and can opt in or out.

Most importantly, the transaction value of using credit cards or rewards systems - what the user actually gets - is not enumerated.

Beyond what others have noted (mainly deferred payment), credit cards offer legal transaction protections: my legal liability for fraud is limited (unlike debit or Zelle transfers), and I can challenge any transaction even later, which gives the vendor an incentive to ensure I'm happy even after they have my money. While reputation provides some incentive for repeat customers, the ability to retract a transaction governs even non-repeats. I would submit this alone has improves quality of service for everyone anywhere credit cards are accepted.

Rewards vary by type. Cash-back rewards reflect the fact that interchange fees were set to recapture initial investments, but servicing costs have plummeting (thanks to computing); governance-wise, it's almost impossible for a "representative" political system to extract a large cost from a small number of powerful agents with vested interests to provide a tiny amount of benefit to a very large numbers of other people. But that's a much more extensive governance issue.

So where does the benefit go? To competition between credit providers, initially as cash-back, and then to tying rebates to future purchases within controlled channels. For airline point systems that give free flights or upgrades, it improves retention, but other forms of rewards would seem to verge on tying, where power in one market is extended into another.

Politically-mediated wealth transfers are a political issue. Economically-mediated wealth transfers should raise market-regulation policy issues, in particular whether the law is inducing or protecting them, and then whether they are good or bad. Tallying that requires not just seeing the money flow, but seeing all the value received or cost exported.

w10-1··on On AI regulation and messaging
> by far the most accurate criticism of AI companies including Anthropic is that we haven’t yet delivered on our big promises to benefit the world. That is totally on us

I don't think it's on them. Proximally, it's not their job to deliver the end-use value; it's their customers' job (if they're doing it, it would be a strong sign of tying). More generally, they operate as a commercial SaaS entity (non-profit or not), and net benefit to the world is not what any company is structured for (even public benefit corporations have mainly indirect effects).

So I think that "criticism" is just a back-handed way of promoting themselves in messianic terms.

IMHO Anthropic, OpenAI et al continue the practices that make people resent companies (US companies): You have to be either connected or nominal (show of diversity) to actually participate in the opportunities or protect yourself from their predations. That's just the nature of knowledge work: you need a ladder of opportunities that start very early to participate, and you'd need to understand subtleties of engagement to see how you're being fleeced as a customer. The divide of rich to poor is not about money but both knowledge and access; it's felt even by long-time experts in the field who are left out or ousted.

w10-1··on DeepSeek Harness developer preview
> Every product relying on "community plugins" for their features implies it works fine the 6 first months, then it's a nightmare

Eclipse has been thriving since 2002 mostly by virtue of being able to coordinate developers via plugin's and a business-friendly license.

They did need to upgrade early plugins into OSGI, and most of the new plugin designs benefit from copying OSGI, et al. The key is SAT solvers for dependencies and namespace separation, not forcing clients into the same dependency version.

But as you suggest, relying on the community is a moral hazard. In Eclipse there were big players willing to fund key use-cases for their own purposes; elsewhere I've seen sufficient monetization of plugins to offer incentives and stability.

I would add that VSCode plugins follow a different development model. While any OSGI/Eclipse plugin can provide an interface, I believe in VSCode you're limited to the API's they give you (and they make a mess of them, so there's more inconsistencies e.g., in LSP support that anyone can enumerate).

w10-1··on Understanding is the new bottleneck
I agree with the problem but not the solutions.

The problem pre-dates LLM's: writing code that "works" but breaks the underlying model. Because it works, it always sounds reasonable and doesn't raise any flags.

Only someone - human or LLM - who holds the model as the standard would see that this working solution breaks the model.

(In theory, the model is to preserve scaling, flexibility or some other systemic feature not immediately invalidated by this working code, but as always the model itself could be bad.)

LLM's are not bad at giving an account of the model; indeed, fighting with the LLM over what the model is can clarify things. But LLM's will happily hold on to a stream of inconsistent statements as their model, so they are not the authority.

w10-1··on Tracking down the 16-year-old WAL-reset SQLite bug
The irony is that the SQLite developers get a support contract iff someone runs off the path in anger and finds an ancient bug. But perhaps that's part of what make it a quality team: devotion thriving without adverse incentives.
w10-1··on Apple Silicon and macOS VMs: Faster LLM Inference with llama.cpp
Related: does anyone have a basis for guessing whether the Neural Accelerators found in M5 Pro+ (accessed by Metal 4) will make their way into the M6 base processors?
w10-1··on OpenAI’s head of ethics leaves less than a year after joining
> what exactly is the business case?

Ethics is also a containment vessel internally.

With an Ethics dept, concerned employees can rest easy knowing someone else is responsible, or if they raise concerns, they have someone to route and advocate their concerns, rather than adding noise to the chain of command. If the dept. ends up ineffective after some years, employees can blame the department, or have renewed hope when personnel turn over. So (slow) turnover is good from this perspective.

w10-1··on Nvidia's Risky Business
Yes and no...

So much of the dotcom bust was essentially: "anything internet will work", but even after it was pretty easy to find different ways to publish web pages or migrate to new forms of social media, video, short-video, etc.

It's orders of magnitude harder to automate work, which is the business value proposition of AI (whether you are displacing worker or adding breakthrough capacity), and it's not clear the LLM hyperscalers will get the value-add from that.

On the consumer side, it might be a race to the bottom: same ad profits, but now you need AI to produce it.

w10-1··on OpenAI's New Device Will Be Hockey Puck-Sized and Cost over $300
Assuming instead of being another phone or voice assistant: It might make sense as a local AI server/proxy expanding to home assistant/TV. If they could claim to need to have device id's to avoid their models being scraped, they could force users to use their device proxy and avoid middlemen (likely by making it cheaper than direct access), and likely save a lot of server time by doing some processing locally on client's dime. Then they could start to play the device-id game with 2-factor authentication, etc. and their stock would track their install base like Comcast STB's.
w10-1··on Everything you do is being recorded
> You don't really explain why being antagonistic to those corporations would be a problem

Yeah, sorry. Yes, corporate activities that reduce competition or corrupt regulators should have their agents and causes ripped out branch to root; they're like diseases which attack the immune system, doubly bad.

But the premise that a corporation or a country that does bad things means you are justified in antagonizing or destroying the entire company or the people is typically enlistment and leverage for protection rackets.

Warring with others is a cycle that doesn't end, and mainly elevates military resource providers. Warring with yourself - attacking your own organs - is self-destructive (which is why enemies stimulate it).

As for here vs there, 20th-century totalitarian systems all stemmed from people wanting to believe promises of a better world, and the promotors being unbound by facts of the present. "Here" is a much more tractable problem, where our medicine (or immune systems) could get better at stemming the diseases we have, like the lack of negative feedback for abusing people (more generally for exporting costs, esp. to the future), or the corruption of discourse that mediates agreements about governance. My goal would be to leave the world more free, peaceful, nature-healthy than I found it. Antagonism doesn't help for that, and side-tracks people into regressive emotions.

w10-1··on Everything you do is being recorded
> we need a government that is antagonistic towards corporations that are abusing us

Please don't. The entire civilized world for all of history and before has been built on the uncomfortable alliance between power/capability and governance/legitimacy. Protection-seeking is the primary support system for tyranny which is happy to induce the chaos they protect you from.

We just need enough negative feedback for any kind of abuse that it stops, because regulations have to be slow to be accurate and thus invite abuse innovation. The main part of that negative feedback is effective competition so people can vote with their feet.

So, ask not for antagonism but for real alternatives.

w10-1··on How I use LLMs to learn complex topics
Gamifying the presentation could make topics more accessible to others. For me the overhead wouldn't help with my own learning. Also I've been burned by just learning things mechanistically (e.g., coding, applying algebraic rules), so I'm leery of learning just by making flashcards or models of the topic.

I find LLM's do great for learning when I ask what are the principles, how the main applications work, what are the key drawbacks, where are the growth plates in the field, etc. - the kind of thing a good advisor points to. Sometimes I have to ask it explicitly to use topological order of topics and show relations, which often highlights the gradient changes in the learning curve. For pruning, it's surprisingly good applying philosophical heuristics - Occam's razor, or Derrida's differance (the difference that makes a difference), etc.

And finally, no learning is effective without problem sets, and for those LLM's at times get me over blocking issues.

The degenerate case is memorizing the glib phrases regurgitated back to me; they're helpful and functional enough to get me into real trouble!

w10-1··on Show HN: Today's cities on a globe of Earth's tectonic past and future
Very nicely done. Plate coloring helped, as did the red highlights for plates pressuring each other.

New for me: I didn't realize eastern russia was at the north pole 220 Mya, and rotated and came south; at the same time, no real China, and India was at the South pole. America's, Africa, Australia didn't change much, but huge changes elsewhere.

w10-1··on App Store Rejection of the Week: Dark Hours
Are those documented anywhere?
w10-1··on Aristotle quotes on virtue, knowledge, and happiness
It's helpful to read the ethics in light of the metaphysics. Beings aren't fixed but actualize their (possibly-growing) capabilities. Happiness lies in friendship, not fame or power or wealth, not because they're not good to have, but because for we humans as intelligent, social, self-aware beings, we can only actualize our craft and self-understanding in the context of others with some understanding of the craft and insight into each other, particularly with respect to building character, one's own and others'.

The first section is simple but profound: all beings strive for the good for themselves, so we have a starting point to address the good of each and possibly all even across fundamental differences.

> he has some pretty messed up and weird views

They do, but by insisting on rationalizing any view rather than leaving it as a religious given, he sets the stage for destroying bias. Remember, Aristotle trained Alexander, and Alexander's military was successful because they embraced rather than destroying the people conquered (if they didn't fight back), and they even lost a fair bit of his Greek ways.

w10-1··on I am retiring from fulltime writing (& pseudonymity) to launch Guardian Angel
I'd be surprised if one's intentions or goals could be trained into LLM's; more likely one would emotionally migrate to the intentions that LLM's can represent because you+they would be more effective.

People already change depending on social context and available prostheses (in Freud's sense of humans as prosthetic gods); since AI is not free (integrations even less so), social divisions along power/wealth lines will increase.

And just as each of us apparently needs a military to defend against the others, so too will we need digital guards for any virtual presence; thus there's no avoiding the who-guards-the-guardians problem.

w10-1··on Dataset: Dead mental health startups, 2000-2026, coded on 18 fields
Post-mortem's are very useful, though often not welcome.

I appreciate the top-line summary and the approach looks promising, but I don't have time to sort through the data or details.

Both are likely common, so don't be disheartened by complaints :)

w10-1··on Twenty Years of Pandoc
unending thanks, as much for pandoc as for the example of a clear design+implementation that lasts.
w10-1··on Don't be a meat proxy
But displacing individual ability is a feature, not a bug.

How else could mind-virus like philosophy or religion or science or political legitimacy transit highly-variable humans for generations?

Religious organizations, governments, corporations all far outlive their founders and maintainers. Heroes at best only stretch or squeeze the dogma or methodology to grow the organization by better fitting to the culture.

And because organizations accumulate control over resources, individuals can have little effect outside them.

We are disposable by design.

w10-1··on UEFA and its national associations will not participate in FIFA competitions
> Do it consciously

Yes

> These words apply to any investment [...] the entire culture has to change

I think that's fatalistic. There are plenty of people and legal structures that limit the influence of investors to reasonable returns, most famously founders retaining control, or the public benefit corporation (like Veeva) or nonprofit (OpenAI as the counter-example showing how even structures are no guarantee). It's all part of setting up the systems to work with available people, within the business leader's ambit.

w10-1··on Three Virtues by Larry Wall
Interesting reminder: Wall's statement of virtues was provocative as a counter to programmers following corporate rules and various software design fads.

But both emphasized personal virtue and reputation in a way that seems quaint in an era where people are schooled to harness systemic forces to their advantage. Even highly flawed people succeed if they manage access to a river of value.

Hypercard lore has it that Jobs asked Atkinson to stop writing fit-to-use software that fits on a floppy and instead write something that requires lots of graphics and has no real functional limits, so Jobs could sell people a mac with an unimaginably large 20MB hard drive.

The free money now likely lies in making use of the oversupply of AI compute once projections fall flat, for uses beyond LLM's. Tomorrow's tech titan will need the following virtues:

Laziness: find the work end-users want to offload instead of protect

Impatience: don't wait for fads or technical swells to start paddling

Hubris: assume you can write new algorithms for AI hardware, and (that) you will find new applications, if you just work at it.

w10-1··on Rethinking legal education in the AI era
These are the right bits to call out.

But the effect is to make success more dependent on networks. (If they are using AI as an excuse to make networks more relevant than skills, it relieves competitive pressure.)

First, a main draw of law journals is the filing cabinet of past exams as remembered by past students, available only to journal members i.e., friends of connected people. Reducing one's ability to remember what was in a lecture only increases the advantage insiders have.

Second, the "Socratic" Method for teaching law is nothing like it is in philosophy; it's more of a weird combination of hide-the-ball and distraction that again ensures people who can see through the clouds are using out-of-band information, and share it only with those they seek to enlist or join.

> all students will be required to engage in an oral discussion of their SRP with their supervising professor, in an in-person setting

I'm surprised anyone could get through otherwise. When did it become legitimate to dump your paper and not answer for it?

w10-1··on Engineering management after the cost of code collapsed
> mistake verbosity for authority

Isn't that the essence of "attention is all you need"?

w10-1··on Who Is America's Homer?
Willa Cather?

Homer, or the Homeric tradition, arose as rootless people displaced by drought and war in ~1,000 BC needed to sell the idea that entertaining stranger guests who had only stories from afar was worth doing, and that they shared common gods (versus the city-specific gods of earlier). This mind-set supported looser city-state alliances and cultures open to trading and ideas, who then outgrew the more xenophobic cities (like the Etruscans). So they semi-accidentally created a legitimizing belief system and thereby stumbled into founder roles.

America as the land of opportunity has flourished by acquiring more land and capturing value better, mainly by diminishing obligations to country and god and fellow humans. Yes, many writers bore witness to the frictions this involved with varying degrees of acuity or tragi-comedy, but no American writer wrote in a way that actually established this American mindset.

I think instead the mindset is perpetuated by each era building achievement heroes, stimulating belief in unbounded personal possibility. American literature is reduced to pointing out the vacuous consequences. As William Carlos-Williams wrote looking into his mirror in middle age, "I am lonely, lonely. I was born to be lonely, I am best so! [...] Who shall say I am not the happy genius of my household?"

If I had a vote, I'd nominate Willa Cather as the writer who portrayed that knowing moment of the powerless, non-achieving people enthralled with the beauty of nature as it was being destroyed. It's not what guides Americans prospectively, but it might best reflect history's perspective and our own retrospectives.

(I should add that I'm glad America has hoisted individuals and done so much to lift the global standards of living and care, even if it was more by greed and necessity than charity.)

w10-1··on Apple targets dozens of OpenAI employees with legal letters
> Apple could take the ChatGPT app down

No. If Apple did that without a basis in their own App Store policies, they'd spawn antitrust cases. OpenAI damages are a droplet compared to the App Store revenues.

> It's insane to cross Apple

... on the question of employee confidentiality. Because the IP landscape is so muddled, they try hard to keep employees happy (most don't leave), and they're very, very good at writing, securing, and enforcing confidentiality agreements.

← PreviousPage 2 of 34Next →