Adding my POV from a former National Security perspective:
Author is 100% on point. The point of a polygraph is three-fold: weeding out the dipshits; exerting power over the powerless; and identifying the valuable assets (typically sociopaths). It does not - cannot - identify liars, deceit, or bad actors on its face (that comes from the manual the author linked). It's not scientific assessment, it's psychological torture.
Would I take a polygraph to reactivate my clearance? Yeah, if I had to. Would I pass? That's up to the examiner, because much like the author I won't tolerate being called a liar, nor will I capitulate to power games. I'll be honest, forthcoming, and cooperative - and if that's not enough to pass, then I don't want to work for you.
This comment or something like it should be at the top, because it's the main point about polygraphing. It's the process, not the answers that matter.
I knew a guy who did security clearance checking for the Three Letter Agencies for many years. He told be that if I ever had to do these interviews, I just need to pick good sounding lies and stick to them. He said it's the ones who try to be honest and introspective who get failed out.
This was all so weird to read about. I guess I just assumed the polygraph was of marginal utility, and you either passed, or you didn't. I didn't realize it was part of a combative interrogation process, even for regular employees.
Not bad, just as misinformed as most folks out there about the process and requirements.
National Security is a PITA, full of cutthroat sociopaths who would eat the SV VC-types for breakfast. That is a compliment, because the work they do is broadly dark and grimly necessary, at least at the levels of global geopolitics a lot of them are expected to operate at. I washed out in contracting for much the same reason this person kept "failing" polygraphs: honesty to the point of external perceptions of naivety. The types who excel in these sectors see folks like us as doormats or tissues, and react poorly when we catch them in the act and demand anything resembling respect because they know we're a threat to the entire establishment if we're allowed to succeed.
The point of polygraphs has always been about control, and folks who resist that sort of control are incidentally highlighting themselves as being uncontrollable to power alone. The books the author links are excellent starting points for understanding the true function of a polygraph, and why more places are outlawing them as a means of trying to diversify a deeply broken and hostile security apparatus by preventing it from being a "blind fools and sociopaths-only" club.
It would seem there's a spectrum of beliefs regarding the people in the CIA, the FBI, in politics, etc. ranging from "They're just like us!" to "They're lizard people (for better or for worse)." In other words, is it the situation or is it the person/self-selection? I self-identify as uninformed about the bigger picture, but my experience working in a federally adjacent sector where all my colleagues are perfectly normal, and yet there is always above us the stench of lizardry in the decisions being made, has me believing in the hypothesis that every bureaucracy is largely staffed with normal people doing the legwork (sometimes very high level, high paying, and highly consequential legwork), and lizards controlling the brain at the management and director levels.
> I washed out in contracting for much the same reason this person kept "failing" polygraphs: honesty to the point of external perceptions of naivety.
I'm curious if you're willing to elaborate on this story. So far in my career I've yet been forced to bend my knee to a lizard, nor become one, but it sounds like you have some experience.
A delightful exercise. Inference and phonetics alone got me back to ~1200 with probably a 90% hit rate. Then it just collapsed under me around 1100.
Honestly not a bad critical thinking exercise in general, for someone with language fluency. Much of it can be “worked out” just through gradual inference and problem-solving, and I’d be curious to see its results as a test for High Schoolers.
Contextual irony aside, this is a big reason why the proposal of leveraging AI agents for workflow processing in lieu of using them to develop fixed software to perform the same functions has always struck me as weird, and of late come across as completely nonsensical.
If you're paying someone else to run the inference for these models, or even to build these models, then you're ultimately relying on their specific preferences for which tools, brands, products, companies, and integrations they prefer, not necessarily what you need or want. If and when they deprecate the model your agentic workflow is built on, you now have to rebuild and re-validate it on whatever the new model is. Even if you go out of your way to run things entirely locally with expensive inference kit and a full security harness to keep things in check, you could spend a lot less just having it vomit up some slopcode that one of your human specialists can validate and massage into perpetual functionality before walling it off on a VM or container somewhere for the next twenty years.
The more you're outsourcing workflows wholesale to these bots, the more you're making yourself vulnerable to the business objectives of whoever hosts and builds those bots. If you're just using it as a slop machine to get you the software you want and that IT can support indefinitely, then you're going to be much better off in the long run.
It's the siren song of the myopically lazy. It's laziness today in exchange for harder work tomorrow, with the wager that tomorrow's harder work will be even lazier thanks to advances in technology.
Whereas I'd self-describe as "strategically lazy". It's building iterable code and repeatable processes today, so I can be lazy far into the future. It's engineering solutions today that are easier to support with lazier efforts tomorrow, regardless of whether things improve or get worse.
Building processes around agents predicated on a specific model is myopically lazy, because you'll be rebuilding and debugging that entire setup next year when your chosen agent is deprecated or retired. Those of us building documented code with agents today, will have an easier time debugging it in the future because the hard work is already done.
Incidentally, we'll also have gainful employment tomorrow by un-fucking agent-based workflows that didn't translate into software when tokens were cheap and subsidized by VCs for market capture purposes.
Love it. Excellent reasoning for subjective decisions that don’t knock the product or solution itself as much as, “not what we specifically needed, and that’s okay”.
Bookmarked for my own infrastructure transformations. Honestly, if Okta could spit out a container or appliance that replaces on-prem ADDCs for LDAP, GPOs, and Kerberos, I’d give them all the money. They’re just so good.
It’s not limited to young people, unfortunately. About fifty years ago, executive leadership became far more visible in the public eye and combative with workers, all to juice share prices for their own compensation bumps. Conglomerates built on monstrous estates of interconnected business lines were gradually gutted and slashed to promote price bumps on shares, at the expense of profitable lines of business.
The net result is a (mostly) American business model predicated on Celebrity C-Suites doing highly visible things while those doing the hard work of creating value are shunted into offices and paid less compared to productivity gains over time. It shouldn’t be a surprise that social media and the internet have supercharged this, especially with groups like YC, Softbank, a16z, and other VCs splashing out Capital on flash over substance, exploitation over business fundamentals, “disruption” over societal benefit and symbiosis.
The net result is a growing schism of resentment by those who do the work towards those who get the credit, glory, and reward, versus those who bask in stardom and truly believe they can replace the perceived entitlement of labor wholesale with an instant gratification machine and somehow survive the resulting societal collapse such a device would bring about.
It's charitable to frame this as resentment towards capital who gets the "credit". I'm sure people would grumble about this regardless, but the real resentment stems from them systematically eroding our ability to afford housing, healthcare, and retirement.
Their unaffordability is only the last straw that will hopefully break the camel's back and create a counter-force.
Normal people generally don't dream to be ultra rich, they just want to enjoy life (and have enough money to do so). But a small percentage is obsessed with money and they obviously invest much more energy into gaining it.
This dynamic means that people don't get paid according to how much value they produce but according to how good they are at negotiating and at maneuvering themselves into positions of power from which they 1) take a bigger cut than they deserve according to real value produced 2) further entrench themselves.
Salary negotiations are a perfect example of divide and conquer - the employer has more information, more runway, more experience negotiating, etc. And on top they negotiate with each employee one by one. Imagine a reverse situation in which the people doing the real positive-sum work sit together on one side of the negotiation table and ask their new assistant (so called "manager") how much he wants to be paid.
But the real issue is ownership. People who don't do any work get paid (if not in money directly, then by being able to sell the company). And they get to pass this "ownership" onto their children who contributed nothing at all.
I am convinced a lot of these runaway feedback loops would be destroyed if ownership of a company was by law distributed among employees according to the amount of time and skill level they worked there.
I like that idea, and I agree. a 10x spread of $ between skill levels, and otherwise by hours of effort and years of tenure. Yes the flight attendant who's worked there for 30 years should have more ownership (and more influence) than an executive who started last month.
I have an idea I've been batting around: mandatory 1% annual tax on public corporations that is expected to be paid in their own stock, and either held in a sovereign wealth fund, or distributed equally to all citizens. This simultaneously dilutes the wealth of the majority owners hold, boosts public savings (tax advantage to holding rather than selling), and makes ordinary people automatically invested in their nation's economy.
> if ownership of a company was by law distributed among employees according to the amount of time and skill level they worked there.
Those are so hard to quantify that I think you'd really have better luck instituting UBI. Both in terms of encoding it into law and getting voter support.
I also want to say, as a market socialist who owns stock, owning stock in your own company is the least diverse investment you can make, except maybe buying a house and then living in it.
And if it's based on time at the company, do I keep the stock when I leave? Am I punished by losing stock if I'm fired? How much of the company is owned by former employees? A lot? None?
If I only own stock while I work there, and I can't sell it, then it's not worth much. It's just a profit-sharing bonus with extra steps.
It's hard to quantify perfectly but we already quantify it imperfectly during salary negotiations. Don't make perfect an enemy of good. We could get a better system _today_ overnight if we just took everyone's salary and used it as weight/skill for distributing ownership per unit of time. We could further improve it by renegotiating on equal footing.
I am not against UBI as a safety net system so that everyone has enough to survive. But instituting UBI before restructuring the ownership system would be actively harmful because, again, we need enough straws to break the camel's back so that people take the time and energy to understand the root causes and oppose them. (Because people's reaction is not linearly proportional to inequality - there's nothing (acceptance/indifference) for a long time, opposition forms only when it's sufficiently bad.)
One large underlying cause of inequality is that we have 2 different reward systems:
a) Fixed money per unit of work (usually per unit of time or per item produced).
b) Ownership which gives full control of the owned structure and therefore the ability to capture the full value produced by it. (Minus money to pay workers but money per person does not scale, ownership per person does.)
These map pretty cleanly to the worker vs owner divide. And this distinction is what we need to erase to erase the class divide.
> do I keep the stock when I leave
Yes, that's the point and this is where it would be better than current co-op systems. Every person's economic input into a collaborative effort is the weight used to divide their ownership. So if you stop working there, you keep your part but it keeps getting smaller relative to the rest as more people keep putting in their work or money.
Money (investment) is a valid economic input and should weight towards ownership. How much? We could use median salary of the country, median salary of the company, or my favorite - divide the investor's total net worth by the number of hours he did worked - this would somewhat erase the advantage rich people would have upon transitioning to this system.
> Am I punished by losing stock if I'm fired?
Interesting question - I don't think so, if you used to contribute positively, you should keep the reward, but you might need to be penalized if you caused harm to the company proportionally to the harm.
> How much of the company is owned by former employees?
That would increase over time up to a plateau as they got old and died. (Ownership should not be heritable.)
---
Regarding diversifying investment - you can do that by either working for many different companies or by using your money to invest into other companies.
Thinking about this as buying and selling shares is IMHO misleading - it's more like re-weighting a distribution. Adding an economic input reduces everyone else's share slightly but since that input (hopefully) leads to more revenue, they will be better off (if they don't think so, they (as owners) can vote against taking the investment).
I am not an economist and I still feel like I am scratching the surface of how the economy works so maybe there are loopholes or degeneracies in this system. I'd like to find them and fix them. And I should probably write a proper blog post about this with diagrams since some of this should be easier to convey with images. What I am proposing is similar to some economic systems (mutualism is one of the closest) but I haven't seen this exact thing around weighted ownership.
Not necessarily. Workers don't want to move into the overclass, they just want to live with dignity. One major theme is that things that seemed very ordinary and attainable a generation ago for ordinary people, like owning a house, now seem out of reach.
Circa 1970 Issac Asimov wrote an essay that started with a personal anecdote about how amazed he was that he could get a thyroidectomy for his Graves Disease for about what he made writing one essay -- regardless of how good or bad it really is today, you're not going to see people express that kind of wonder and gratitude about it today.
but I think the real working class stance is that you want protection from economic shocks more than "participation", "ownership", "a seat at the table", "upside", etc. This might be a selfish and even antisocial thing to ask for over 80 years near the start of the second millennium, but I think it would sell if it was on offer. It's not on offer very much because it's expensive.
One could make the case that what we really need is downward mobility. Like what would have happened if Epstein had been shot down the first time or if Larry Summers had "failed down" instead of "failing up?" My experience is that most legacy admissions are just fine but some of them can't test their way out of a paper bag and that's why we need a test requirement.
> Workers don't want to move into the overclass, they just want to live with dignity.
Got it in one. Would I like to travel First Class and stay in fancy hotels? Sure, but I’d much rather have a house that I can improve to meet my needs instead. Would I like a fancy luxury car with all the trimmings over my sixteen-year-old Honda? Absolutely, but the latter is paid off and gets us around just fine. Would I like that spiffy Hasselblad X2D and some lenses? You betcha, but I’d rather take a proper holiday for the first time in fifteen years instead of buying another thing.
The problem is that society at present isn’t organized to prioritize necessities like shelter and healthcare, favoring wealth extraction and exploitation instead. Workers don’t want megayachts and hypercars and butlers, we just want to live more than we work.
I love the idea of "downward mobility". In particular over the past 30 years we've created a new class of ultra-ultra-rich with even more wealth than the robber barons of the gilded age had, and we need to figure out how to dismantle that entire class. A puny 3% wealth tax would take over 100 years to knock them down, and that's presuming that their wealth is static and not growing at a rate much greater than 3%.
Can you recommend some positive sum games for kidults who have a hard time getting angry ---even when short on gratitude--- but would like to keep it that way* :)?
You clearly don't know what the term upward mobility means. It doesn't necessarily mean moving from one class to another - though that WOULD be included within its scope, however extraordinary an example it may be.
It can mean moving within a class.
Surely most people want to better their station. To argue against that is insane and counter to every observable fact about human nature.
It can, but it's not how it's used most of the time, so kind of a pedantic distinction.
And many do not even want to "move within a class" that much. They'd be satisfied to keep their job and retain the same constant purchasing power and ability to buy food, feed family, pay rent/morgage, year after year.
Jocks becoming nerds or vice versa or both becoming hipsters, are examples of intraclass mobility.
There's also cultural mobility which is different from economic mobility.
All are exemplified by reddit's Ohanian marrying one of the Williams sisters and thus having a either higher or lower social status, than either Ghislaine or Larry Summers, quite independently of how much cash they each have in the bank
I wonder if popalchemist would count the cultural station as something worth improving apart of the economic one
> Those who control capital use their political and economic power to systematically enrich themselves at the expense of those who actually perform useful labor
Huh, I think I read a book about that once. I forget who wrote it. Carl something, I think?
Yep, that played a significant role in shaping how things turned out. We want a single source to blame, but rarely does history present us with such a neat villain (though god, Reagan comes so close to being one, at least for the specific issues important to me).
Understanding the interconnectedness of systems beyond your own realm of expertise is how you learn what needs to be done to fix issues - and avoid falling for snake oil “silver bullets”/“one weird trick” populist positions.
>The net result is a growing schism of resentment by those who do the work towards those who get the credit, glory, and reward, versus those who bask in stardom and truly believe they can replace the perceived entitlement of labor wholesale with an instant gratification machine and somehow survive the resulting societal collapse such a device would bring about.
Naturally, unmentioned are those shut out of reasonable opportunities for meaningful productivity, regardless of technical potential (but largely in line with (lack of) social capital). A few years of this maybe encourages an entrepreneurial spirit. Two decades is quite convincing that there's no place for them in the current order.
The upwardly-mobile opportunity hoarders need to understand, much as the wealth hoarders ought to, that the whole thing falls apart without buy-in from the "losers".
I’ll second this. Google’s investment in underlying accelerators is the big differentiator here, along with their existing datacenter footprint.
Everyone else has to build infrastructure. Google just had to build a single part, really, and already had the software footprint to shove it everywhere - and the advertising data to deliver features that folks actually wanted, but could also be monetized.
I was thinking about that (I definitely agree with you on the software and data angle).
But when you think about it it's actually a bit more complex. Right now (eg) OpenAI buys GPUs from (eg) NVidia, who buys HBM from Samsung and fabs the card on TSMC.
Google instead designs the chip, with I assume a significant amount of assistance of Broadcom - at least in terms of manufacturing, who then buys the HBM from the same supplier(s) and fabs the card with TSMC.
So I'm not entirely sure if the margin savings are that huge. I assume Broadcom charges a fair bit to manage the manufacturing process on behalf of Google. Almost certainly a lot less than NVidia would charge in terms of gross profit margins, but Google also has to pay for a lot of engineers to do the work that would be done in NVidia.
No doubt it is a saving overall - otherwise they wouldn't do it. But I wonder how dramatic it is.
Obviously Google has significant upside in the ability to customise their chips exactly how they want them, but NVidia (and to a lesser extent) AMD probably can source more customer workflows/issues from their broader set of clients.
I think "Google makes its own TPUs" makes a lot of people think that the entire operation in house, but in reality they're just doing more design work than the other players. There's still a lot of margin "leaking" through Broadcom, memory suppliers and TSMC so I wonder how dramatic it is really is
My take is it's the inference efficiency. It's one thing to have a huge GPU cluster for training, but come inference time you don't need nearly so much. Having the TPU (and models purpose built for TPU) allows for best cost in serving at hyperscale.
Yes potentially - but the OG TPUs were actually very poorly suited for LLM usage - designed for far smaller models with more parallelism in execution.
They've obviously adapted the design but it's a risk optimising in hardware like that - if there is another model architecture jump the risk of having a narrow specialised set of hardware means you can't generalise enough.
Prefill has a lot of parallelism, and so does decode with a larger context (very common with agentic tasks). People like to say "old inference chips are no good for LLM use" but that's not really true.
NVidia is operating with what, 70% gross margin? That’s what Google saves. Plus, Broadcom may be in for the design but I’m not sure they’re involved in the manufacturing of TPUs.
Yeah this is a bummer. If it goes south everyone in power will also have perfect hindsight and say they saw it coming because obviously you shouldn't have this much built on such a small footprint. And yet...
> Yeah this is a bummer. If it goes south everyone in power will also have perfect hindsight and say they saw it coming because obviously you shouldn't have this much built on such a small footprint. And yet...
It'll be true, everyone does see it coming (just like with rare earth minerals). But the market-infected Western society doesn't have the maturity to do anything about it. Businesses won't because they're expected to optimize for short-term financial returns, government won't because it's hobbled because biases against it (e.g. any failure becomes a political embarrassment, and there's a lot of pressure to stay out of areas where businesses operate and not interfere with businesses).
America needs a lot more strategic government control of the economy, to kick businesses out of their short-term shareholder-focused thinking. If it can't manage that, it will decline into irrelevance.
When USSR fell there was a lot of talk about how it was meant to be since the US system is the best of all the terrible systems. It deserved to win and the USSR system deserved to die.
>America needs a lot more strategic government control of the economy, to kick businesses out of their short-term shareholder-focused thinking. If it can't manage that, it will decline into irrelevance.
If it is meant to be then its meant to be. If the US decides to cling to its old system and it fails well then we would know that it wasn't the best system after all. Humankind will keep moving forward even if it means that another continent controls the show.
Just by removing nvidia's profits Google gets their TPUs for like 25-30% of what they cost to get them from nvidia, assuming similar cost structures. Google's cost structures are probably higher than nvidia's so realistically theyre probably paying around 50% of nvidia charges, but thats still billions of dollars a year and allows them to create a product that is tailor made for their needs.
While I think Gemini is the worst of the three big competitors, Waymo is an superb example of this talent. Kudos to Google engineers for producing so many diamonds despite producing many terrible flops over the years. We might find out their system of organization was the best one after all.
I still believe this is the right - and inevitable - path for AI, especially as I use more premium AI tooling and evaluate its utility (I’m still a societal doomer on it, but even I gotta admit its coding abilities are incredible to behold, albeit lacking in quality).
Everyone in Capital wants the perpetual rent-extraction model of API calls and subscription fees, which makes sense given how well it worked in the SaaS boom. However, as Taalas points out, new innovations often scale in consumption closer to the point of service rather than monopolized centers, and I expect AI to be no different. When it’s being used sparsely for odd prompts or agentically to produce larger outputs, having local (or near-local) inferencing is the inevitable end goal: if a model like Qwen or Llama can output something similar to Opus or Codex running on an affordable accelerator at home or in the office server, then why bother with the subscription fees or API bills? That compounds when technical folks (hi!) point out that any process done agentically can instead just be output as software for infinite repetition in lieu of subscriptions and maintained indefinitely by existing technical talent and the same accelerator you bought with CapEx, rather than a fleet of pricey AI seats with OpEx.
The big push seems to be building processes dependent upon recurring revenue streams, but I’m gradually seeing more and more folks work the slop machines for the output they want and then put it away or cancel their sub. I think Taalas - conceptually, anyway - is on to something.
This is a blessing for us Dynamic DNS folks whose DNS providers demand a static IP for changes to come from (i.e., Namecheap). In theory, it means we can set this up once (or on a schedule that works for our needs), and trust that renewals will happen without continued maintenance or involvement.
Eager to give this a try as I modernize the homelab.
Author is 100% on point. The point of a polygraph is three-fold: weeding out the dipshits; exerting power over the powerless; and identifying the valuable assets (typically sociopaths). It does not - cannot - identify liars, deceit, or bad actors on its face (that comes from the manual the author linked). It's not scientific assessment, it's psychological torture.
Would I take a polygraph to reactivate my clearance? Yeah, if I had to. Would I pass? That's up to the examiner, because much like the author I won't tolerate being called a liar, nor will I capitulate to power games. I'll be honest, forthcoming, and cooperative - and if that's not enough to pass, then I don't want to work for you.
reply