cornholio

> They wouldn’t “handle” it, they’d have some third party do their dirty work.

Without proof, that's just an edgelord conspiracy theory.

Police are not the Borg, perfectly coordinated in their evilness, all law enforcement agencies have internal power structures and strife, rivalries, jealousy, old conflicts. The fact that some action, such as planting evidence leading to a conviction, is punishable with long prison sentences, is not something the corrupt can simply afford to ignore, while giving their internal foes mortal leverage against them.

For example, if Kash Patel receives an order from his handlers to plant child porn on some political target, that outcome might happen or not, but what you can be pretty damn sure is that all those involved will be aware of the risks and will try their best to stay out of it, or, if coerced, do it covertly so as to minimize the extreme risks they face.

The point was not that FBI are a bunch of angels, but that the undeniable risks involved by such a move seem completely unnecessary - the FBI has for years been weaponized against overseas copyright infringers, openly and legally.

It's unlikely law enforcement would take the risk to handle CSAM just to make a case against a Russian pirate, jeopardizing their careers and freedom, when the copyright case is pretty strong already.

These are the doings of one of the myriad freelance "intelectual rights enforcement agents", which are paid on success and employed by some large media organization. Another possibility is that a single aggrieved individual who found themselves doxed or their criminal conviction archived etc. took action after failing to enforce their so called "right to be forgotten".

Unfortunately, archive.is operating model is uniquely vulnerable to such false flag attacks.

Sure, if the common denominator user is at least as savvy as the entire marketing and strategy departments of these trillion dollar companies, then sure, users will identify products that are not designed according to their best interests and will then perfectly coordinate their purchases so that such products fail in the marketplace. Sure.

> no longer building software for their users but as vehicles to push some agenda

All companies push an agenda all the time, and their agenda always is: market dominance, profitability, monopoly and rent extraction, rinse and repeat into other markets, power maximization for their owners and executives.

The freak stampede of all these tech giants to shove AI down everybody's throat just shows that they perceive the technology as having huge potential to advance the above agenda, for themselves, or for their competitors at their detriment.

Great passion for the subject, definitely doesn't get discouraged by their less than perfect command of English and didn't use an LLM to butcher the text's authentic character.

I find that in my own writing I no longer strive for perfect grammar and polish since nowadays it actually cheapens the end result, everybody has perfect grammar today.

They don't care about the resell value since they don't earn a commission on those sales.

The point is that, for as long as items can be transferred in game, they are always convertible to cash in the real world. Inserting artificial friction inside the game to increase scarcity, such as limiting convertibility of items, will drive those trades away from the game economy and into the third party ecosystem where the dollar rules supreme as the super-convertible means of exchange. So you have an induced scarcity that in effect drives third party profits.

By increasing in-game convertibility, the trades are directed to other in game assets that are a just a proxy for loot boxes, i.e money in Valve's accounts. So prices crashing in the third party market signal that players have a cheaper and more direct route to acquire them - give the money to Valve - which also generates the supply of new rare items as those loot boxes are opened.

It's a smart economic move.

Buy that doesn't mean the prices will stay low, since they can always control the overall scarcity, or add new, rarer and more exclusive items. The total amount of money they extract from "kids" is ultimately linked to their ability and willingness to pay.

The purpose of the update is certainly not to reduce the cost of these items, but to better position Valve to earn this revenue steam, as opposed to third party scalpers. Looks like it's working.

The trait of caring and providing for other injured or vulnerable members of the group is widespread in the animal kingdom, there is even a name for it, allocare. Since social groups usually form based on family kinship, it can be explained on selfish gene dynamics.

So the argument was always gradual, not that social care is unique to civilization, but that it happens to an extent (such as the very long recovery period and food resources required to heal a femur) that we can arbitrarily call "civilization".

On the other hand, you could stretch that Scotsman in the opposite direction: do we really provide enough care to other people to the point we are different from animals and can claim ourselves truly civilized?

We've got our best spike sharpeners working on improvements.

I wonder when we will see generative AI codecs in production. The concept seems simple enough, the encoder knows the exact model the decoder will use to generate the final image starting from a handful of pixels, and optimizes towards lowest bitrate and minimum subjective quality loss, for example, by letting the decoder generate a random human face in the crowd, or give it more data in that area to steer it towards the face of the team maskot, as the case may be.

At the absolute compression limit, it's no longer video, but a machine description of the scene conceptually equivalent to a textual script.

The economic notion is called marginal profitability. Better sales are a good thing if the marginal profit is positive, ie, each extra unit sold still increases the overall profit, so in your example it's still profitable if the new model brings $1.5 profit per unit, and you stop only when the marginal profit per unit turns negative.

In tech the model is often misleading, since the large investments to improve the product are not just a question of current profitability, but an existential need. Your existing product line is rapidly becoming obsolete and even if it's profitable today, it won't be for too long. History is full of cautionary tales of companies that hamstrung innovation to not compete against their cash cows, only to be slaughtered by their competition next sales season. One more to the pile.

Sure, but the herd can still be right from time to time. While it's impossible to accurately predict the future course of AI progress, we can at least agree that the rapid take-off scenario is at least plausible, and that, if it happens, it logically leads to dramatic social and political changes unprecedented in the last centuries; in that scenario, AI-empowered human political operators will quickly become dangerous to humanity and they pose a far greater and earlier threat than the comic book scenario where unaligned AI takes over. We had unaligned HI from the very beginning.

Because it lowers the threshold for a total informational compromise attack from "exfiltrate 34PB of data from secure govt infrastructure" down to "exfiltrate 100KB of key material". You can get that out over a few days just by pulsing any LED visible from outside an air-gapped facility.

Interesting. The drawbacks you describe seem to be limitations of the sensor and image processing technology available almost 30 years ago.

For example, my Epson inkjet printer can do about 10ppm at it's lowest print quality, so it can mechanically move and scan a page against the printhead every 6 seconds; a 1Mpx sensor with a 60Hz frame rate will generate 360Mpx in 6 seconds. Even if you throw away 50% of the data (overlap areas, next page load, motion or optical blur at the edges etc.), that's still enough data for a ~1400 dpi raw resolution of an A4 page at the fastest speed. If you are willing to go slower, the resolutions the system could achieve seem outside the range of any flatbed.

Of course, you would need e very beefy image processor to handle the multiple Gbps raw video data and process in real time down to the final scan image, but the actual corrections seem very achievable with modern algorithms.

Outside of the cost of the image processor, another showstopper I can see is motion blur on the sensor, stopping the heavy printhead from its inertia, so that you can have a still image, will kill your total scan speed. But perhaps you can just pulse the LED, or a gas discharge lamp, and impress the sensor with near instantaneous flashes of light.

I wonder why the system didn't caught on and why it's not used today by manufacturers of multi-functional printers. Seems like a huge opportunity to use the existing paper handling mechanism - with an autofeeder, a feature most flatbeds lack! - and get a more compact device.

The entire device consists of a single, cheap CMOS image sensor, a lens focused at a fixed distance and a RGB led. Everything else, stitching the resulting scanbands, correcting for mechanical and optical distortions, etc. is all in software. The native optical resolution you could expect from, say, a 1080x720 px sensor would be something like 2400 DPI.

The only downside i can think is that you can't scan IDs, passports etc. and the location near the inkjet head tends to get dirty.

So all those CEOs and CTOs are wrong, and you along with the skeptic camp see the true nature of AI.

But let's assume for the sake of argument that you are wrong. What are the logical implications? Can your worldview allow for the possibility that you might be wrong?

The history of all hitherto existing society is the history of power struggles.

That's the entire economic premise of AI: labor automation. The end game is more returns to financial capital and less (or no) returns to human capital.

The global employment in software development and adjacent is in the tens of millions. To say the impact of AI code automation will be, at max, a rounding error of just 1-2% of that is just silly; currently, the junior pipeline is almost frozen in the global north, entire batches of graduates can't find jobs in tech.

Sure, the financial math over 30 years does not follow elementary arithmetic, and if the development hits a wall tomorrow they will have trouble recovering the investment just from code automation tools.

But this is a clearly nonsense scenario, the tech is rapidly expanding to other fields that have obvious potential to automate. This is not a pie-in the sky future technology yet to be invented, it's obvious productization of latent capability, similar to the early internet days. There might be some overshoots but the latent potential is all there, the AI investments are looking to be the first movers in that enormously lucrative space and take, what seem to me, reasonable financial risks in light of the rewards.

My claim is not that AGI will soon be available, but that applying existing frontier models on the entire economy, in the form of mature, yet to be developed products, will easily generate disruption that has a present value in the trillions.

I don't think I follow. 1 trillion total investment divided by 100k yields 10 million man years, or 300k man-careers.

Just the disruption we can already see in the software industry are easily of that magnitude.

No, that's not the claim. The claim is that we will create a hypothetical LLM that, when tasked with a problem at the scientific frontier of molecular biology will, about 10% of the time, correctly reason about existing literature and reach conclusions that are valid or plausible to similar experts in the field.

Let's say you run that LLM one million times and get 100.000 valid reasoning chains. Let's say among them are variations on 1000 fundamentally new approaches and ideas, and out of those, you can actually synthesize in the laboratory 200 new candidate compounds, and out of those, 10 substance show strong in-vitro response, and then one of those completely cures some cancerous mice.

There you go, you have substantially automated the intellectual work of cancer research and you have one very promising compound you can start phase 1 trials that you didn't have before AI, and all without any AGI.

It's silly to say that the only objective that will vindicate AI investments is AGI.

Current batch of deep learning models are fundamentally a technology for labor automation. This is immensely useful in itself, without the need to do AGI. The Sora2 capabilities are absolutely wild (see a great example here of what non-professional users are already able to create with it: https://www.youtube.com/watch?v=HXp8_w3XzgU )

So only looking at video capabilities, or at coding capabilities, it's already ready to automate and upend industries worth trillions in the long run.

The emerging reasoning capabilities are very promising, able to generate new theories and make scientific experiments in easy to test fields, such as in vitro drug creation. It doesn't matter if the LLM hallucinates 90% of the time, if it correctly reasons a single time and it can create even a single new cancer drug that passes the test.

These are all examples of massive, massive economic disruption by automating intellectual labor, that don't require strict AGI capabilities.

30PB / 100Gbps comes down to about a month, 4 links would give you a week, so that seems pretty quite acceptable for a training run, especially since you can overlap the initial loading of the array with the first training, i.e train as data becomes available.

It goes without saying any data pre-processing needs to be done before writing, at the storage site, or on the training GPUs.

> nothing making passing organisms more "efficient". You can clearly observe it with flightless birds on small islands.

I'm not sure I follow this argument. Flightless birds happen on small island when those habitats are free of significant predators, for example after some extinction event or due to volcanic origin. Birds can still reach the islands, at least for some geological period, and evolutionary pressure favors some of them to grow in mass, bone density etc. to better control the available land food sources such as seeds, grasses, insects and other small critters.

Flight is not a useful trait in this environment and is expensive to maintain; a flightless bird will always win a fight against a flying bird for the same land source of food, since flight requires elongated and brittle bones, a low mass of muscles optimized for endurance etc. So flight is selected out of the land feeding niche very much for "efficiency" reasons.

The key words here are "a basic" level of housing. A house in the most exclusive area of town will always be an investment (not necessarily a good one), because it primarily offers exclusion of other poorer members of the society from your surroundings, not habitation for yourself. It can't be affordable by definition.

But a basic level of housing is a human right, because it's a prerequisite for maintaining your humanity, ditto for healthcare.

> Adding "resilience" usually adds to the per-unit cost.

In this case, the cost is much less than a dollar (say, a varistor that blows the existing fuse) and it prevents a catastrophic failure.

Panels are not standardized and are themselves series-parallel constructions with wildly diverse specs.

It probably depends a lot on the ecological niche. For example, for a herbivore species, the marginal benefit from the high energy thinking mutation is low, it took millions of years of thinking in humans before we could reach the levels where agriculture was possible, and it was invented by people with a similar biological intelligence to modern humans.

Whereas for a social hunter species, there are ample strategic and game theoretical benefits from even primitive thinking, and the dividend is enormous, since all of a sudden, with proper planing and coordination, you can hunt much larger animals than yourself, which can sustain the community for a very long time and at high energy levels which further enable selection for thinking.

It's only after such large pray is depleted and there is no space for migration, because you are surrounded by other packs of competitive thinkers, that the need to control plant growth becomes imperative.

This is clearly not true: once thinking emerges in a species, it can be applied for an infinite number of situations that can't be codified into instincts because they are unique. Each battle site is unique, it has a certain topology, each battle involves a particular snapshot of men with certain equipment, gear, training and morale etc. The commander who can reason generally about battles will win on average against the commander that can simply "feel" the right moment to attack, if that intuition is just evolutionary inheritance from a long line of warriors who survived very different battles in their time.

So the advantage even against members of your own species is substantial and quickly leads to reinforcing and spreading of the thinking gene.

Social ties correlate with all kinds of beneficial traits, outcomes and privileges, in a very complex and bidirectional causal relation: sociable people have much better economic and career prospects, healthy middle class people have the opportunity, time and resources to engage in social activities, raising a family is a high energy activity not everybody can afford that basically "generates" a substantial web of social links, which in turn support the person in their gray years, and so on and so forth.

So the finding here is that healthy, wealthy people with a support network age just like healthy, wealthy people with a support network.