Their privacy policy for normie subscribers says in plain English they use your Personal Data for research. I think it’s pretty unreasonable to use the service and expect otherwise.
My doctor's privacy policy is a bit more abusive than OpenAI's. It exists mostly because of a $%^&&* legal framework rather than malice, and I've grown accustomed to "if I don't want to die then I sign away these rights." Despite my having theoretically signed my soul away, my doctor isn't selling personal information to my exes or to life insurance companies (though they could in the US; that extremely personal information is no longer mine). OpenAI is engaging in the "technically legal maybe we'll see but obviously unintended" side of this transaction, and maybe that works out for them, but I wouldn't personally choose to be a shill for "it's unreasonble to expect somebody with 'legal' permission to do something other than the maximum 'legally' permitted" if I were in your shoes.
I have the setting turned on in Gemini Pro even though I work on proprietary code because 1. the setting allows for some (very limited) "memory", and 2. I consider my source code almost public even when it's not open source because I don't work on programs that involve extremely high level of know how or proprietary algorithms. It's mostly CRUD that can be copied in a myriad of ways, whether people use my methods or other methods.
If Gemini can improve based on my code and sessions (maybe doubtful but who knows) and others can benefit from it, that would be a welcome side-effect.
Depends on your profession? I sometimes work on open source code as part of my professional duties. Nothing that goes on there is necessary to keep private.
Aren't business consulting firms even worse? They are explicitly for business and there are known cases where they shared confidential information of one of their customers with another one.
> You'd think theft would still be illegal regardless of what a privacy policy says.
I wish that were true, but I live in the United States and it is 2026.
The President of the United States rug-pulls memecoin crypto and regularly pardons people like Paul Walczak (who was convicted of massive payroll fraud) in exchange for large donations.
I wouldn't make any assumptions about what is considered theft anymore, at least not when it is being committed by people who have enough money to be above the law.
You could probably train this out. I don’t think you need to develop elaborate filters. It doesn’t seem like that big a hill to climb if it’s important to people.
That's why this paper is important - it shows it isn't trained out. Leaving no other information in the model makes it clear what the biases are, and that the model is willing to make a biased decision. If you give it other unbiased criteria as well the bias may still easily remain but not be as clear.
Not sure it’s that strong. The prompt gives the presumption that this matters. Not necessarily a training issue vs the prompts being poorly written and the results being inherent in the bias they carry
Eh, I tried in 2017 with BA when their system went down and received nothing. There was magically a Black Lives Matter protest at Gatwick (or maybe LCY, can’t remember) on the same day, which isn’t covered apparently.
Almost certainly not. The legislation says: "An operating air carrier shall not be obliged to pay compensation [...], if it can prove that the cancellation is caused by extraordinary circumstances which could not have been avoided even if all reasonable measures had been taken."
That gets them off the hook for even second order effects. No harm throwing in a claim though, there's no penalty for having a claim turned down.
Isn’t their whole thing supposed to be spying on foreigners? They seem to be quite successful. There aren’t that many exchanges [1]. Could probably manage with cash, guns, and some know-how.
If you just look at the largest 4 of those, you'd have 100Tbps of traffic to monitor, with an average throughput of roughly half of that.
That's ~540PB ((50 Tbps / 8 bits) * 86400 seconds/day) of traffic a day with just those four. Add in the rest and you're likely talking ~Exabytes of data each day. And that has to all be processed on site.
If someone wants to argue that the NSA is in these facilities I'd be 100% onboard. But inspecting it all would be nearly impossible, let alone capturing it all and sending it back to some datacenter somewhere, which is a physical impossibility.
That's nothing a rack full of fast switches can't handle. A rack full of fast switches already does handle it - where do you think the original copy came from?
They will get a copy of the whole feed, but not store all of it - they will have heuristics for selecting interesting traffic.
Switches handle data at far faster rate than any hardware can actually inspect it, store it, process it, etc.
But yeah, just a rack of "fast switches" is all it takes to route hundreds of petabytes of data each day. You should let the data center operators know. They'd save billions.
Switches do inspect it. They also have a feature designed for wiretapping, which copies a percentage of traffic to another port. They may have a feature to copy 100% of traffic matching a certain filter. Managed switch ASICs have this feature even though it's usually not exposed in the CLI.
Vacuous, like calling a V8 a “next piston firing predictor” because engines are designed so that one piston sets up the next in the firing order and technically there’s some nonzero probability any piston can (mis)fire next. It’s missing two pieces:
1. Useful work that has been done (the previously generated token sequence :: the mechanical work already accomplished)
2. The role of structure in relation to the application (post-training :: other components like crankshaft etc)
A V8 does not "predict" the firing of the next piston, it triggers the firing of the next piston at a precisely controlled time with a spark plug (or a fuel injection nozzle in the case of a diesel engine).
The output of the LLM is literally a probability distribution of what the most likely next token is.
> A V8 does not "predict" the firing of the next piston
It kind of does, though. In a gasoline engine you need to spark the combustion in advance of the piston reaching top dead-center to ignite the fuel early enough that it is able to provide downward pressure on the piston as it rolls over top dead-center. The amount of advance required changes with RPM, fuel octane, etc.
Start of delivery timing in a diesel is similar. You have to do it sufficiently far in advance to account for compressibility of the injection lines, fuel burn rate, etc as a function of RPM. A mechanical governor on an injection pump has a timing advance device built in. Electronically governed injection pumps, or modern common rail systems, do that in software.
So mechanically, engines kind of "predict" the next combustion event. Even moreso when you consider a modern ECU, which may be working at nanosecond resolution to time multiple injection events per cycle. To do this at such a resolution it will have to send signals to components based on a predictive model derived from "past" sensor data. E.g. it needs to act ahead of time to account for electrical and mechanical delays in the system.
Interestingly drinking more water was associated with worse outcomes. I would hesitate to point to dehydration without evidence, as a harmful remedy is worse than doing nothing.
Moonshine consumption could probably be investigated further as it carries explanatory power.
I dunno man. Look at the chart, look at the graph, look at the progress, but I live on Market, and it sucks now. Cop cars parked at night with lights flashing since there’s such little traffic. Plus a lady had her head popped open a few weeks back by a pro driver.
I’d be more comfortable making it a Waymo superartery. I trust those things more than the furniture mart box trucks because one slip and you’re gone.
I mean looking at the state of that conflict… the low-cost drone thing doesn’t look that great. Maybe other countries will want to go with this kind of attritable aircraft concept instead.
I think they meant the kind of warfare cheap drones encourage. I believe they implied other means may result in quicker conflict resolution.
I think the problem is, apart from deploying nuclear weapons, air-based warfare campaigns have never really been all that effective at achieving strategic goals. Bombing stuff always seems promising, but in reality consequences are hard to predict due to complexity and results tend to be underwhelming.
For the same reason, cheap drones favor the defender IMO. A typical country can sustain quite a bit of damage on home soil, without losing will to fight and popular support. On the other hand, the attacking carrier on their coast line, becomes a singular target instrumental to the aggressor's war efforts. On the front, cheap drones freeze the contact line, which of course also benefits the defender.
Ukraine is very successful at keeping front lines relatively stable, despite man power shortages, due to drones, but neither Ukraine, nor Russia are very successful with their deep strike drone campaigns. I mean stuff is burning all over, but it changes little for ending the war so far. Train tracks are quickly back in operation, refineries repaired, stocks replenished, behavior patterns changed. Population experiencing reverse hedonistic adaptation to the new shitty normal. No models to predict resilience. Collapse may happen instantly or not at all.
Also pretty evident with the US/Israel aggression against Iran right now. It somehow always takes a lot more than expected to break your enemy in the deep.
> For the same reason, cheap drones favor the defender IMO.
Disagree completely 1,000,000%
For a few thousand dollars, you could attack a major highway or rail network or port a few times. The cost to defend would be astronomical compared to the cost of the attack.
Defense against cheap drones is the real drone cost, not the drone itself.
I meant defender, as in country being attacked. As in not the the aggressor.
In the concrete instance of AA defense you are correct. You basically can't prevent drone damage actively, since every AA system can be easily overwhelmed and costs are very asymmetrical. However, the damage taken does not change much, strategically. Highways, rail tracks and the ports are quickly operational again. That's my point. See war in Ukraine. Nearly every refinery in Russia has been hit by drones. Complex structures, yet most of them are operational and sustained income damage is comparatively low. Of course that's expensive for Russia, but most of that economic damage is offloaded to the future. Same the other way around. Russia is successfully bombing Ukraine every day, yet the front line has hardly changed in years.
Also, the immediate AA defense perspective is a momentary picture. In the future we will see more underground/protected infrastructure immune to drones, as well as more symmetrical defense options (cheap drone on drone systems, lasers, ...). Cheap drones will not become bunker busters any time soon. Doctrines will also likely adapt more rapid initial escalation since nobody wants the prolonged attritional warfare expected with cheap drones.
This depends on the threat scenario. Sure, in Belgium or the US you won't see defense factories going underground any time soon. But Poland? Doing so already proved a valuable investment for Iran - even facing an enemy with an abundance of fancy bunker busters. Ukraine also started to build energy infrastructure and weapon production sites underground, and even Russia finally started building protected shelters for their military aircraft. I am pretty sure the US will also reconsider their arrogant assumptions of high tech superiority going forward, when it comes to their bases in the middle east, where they got very, very expensive stuff destroyed by Shaheds oversaturating Patriot systems. You see, deep underground cities and bunkers withstanding nuclear strikes are not required. You only need a bit of concrete and dirt to withstand medium yield flying bombs arriving with low kinetic energy (unlike ballistic missiles). Hobbit style housing. A cheap drone will never be able to dig itself into its target, since cheap and fast are incompatible.
That said, of course some important components can't really be protected passively. E.g. comms and sigint will always have expensive, hard to replace antenna arrays exposed. But narrowing vulnerability down to those assets means you can focus scalable air defense and EW measures there.
Again, this benefits the country on the defense, especially against distant enemies, since you can build bunkers etc. at home, but can't really bring them along. And bunkers are one time investments, theoretically good for centuries (WWII/Soviet bunkers are still all over the place in Europe). I think, it would be viable for a peaceful country to almost entirely invest military funds in bunkers, drones, drone comms and crews/AI, and possibly nukes, to discourage any aggressor, while minimizing peacetime running costs . Additionally, highly decentralized renewable energy like solar and wind makes that sectors really hard to disrupt. Especially individual solar installations, battery storage and electric cars potentially enables quasi autarky for large portions of the population. Mighty "petrol station" Russia, got fucked by their reliance on ICE. Solar and wind will work even in nuclear wastelands, and replacing that solar panel is still cheaper than the cheap drone, which damaged it.
Tl;dr: You can restructure military doctrine and defense focus to make bunkers an easy sell to your citizens. In fact, going any other way should raise questions about aggressive ambitions.
Ty for the insight - indeed, i had been thinking about full nuclear bunker, which you point out is overblown. That said, complacency (~entropy) tends to win over time.
Ukrainian yong adult in 2051: ouch!
Old Ukrainian: first time in a bunker? mind your head.
> Highways, rail tracks and the ports are quickly operational again. That's my point.
Have drones drop explosives on the LA freeways for a week and see what happens.
It does a tiny amount of direct damage.
But the fallout, politically, economically, etc would be immense.
We're talking about billions spent across the country to try and harden critical industries and transportation, harden buildings, set up various types of drone point defense, funneled in as fast as it can be done least a politician not be reelected.
I'm sure somebody evil enough and creative enough could find a way to get the president to deploy the national guard across the country for an extended period. That's not at all cheap.
Just think about 9/11 for a sec. That could still happen today with a handful of drones.
> I believe they implied other means may result in quicker conflict resolution.
Making wars unending quagmires is an excellent form of preventing them. Had Putin known how bad it would turn out, I doubt an invasion would be his option of ensuring long term access to Sevastopol.
Yes, I also think cheap drones are somewhat of a pacifying development, similar to nuclear bombs, but available to pretty much any nation. (Although, I think ironically nuclear strikes become more likely now, because of drone asymmetry...)
Then again, the US learned nothing from Ukraine and still attacked Iran - the Shahed country. Hopefully, China is a bit more considerate about Taiwan.
reply