> Are you familiar with the Al-Ahli hospital incident in Gaza?
I am not
> Those of us who paid attention learned to not rush to conclusions, and never, ever trust social media or the western press to overcome or even understand information warfare.
Since you highlight western press can't be trusted to overcome / understand information warfare, would you care to provide some write-ups detailing the viewpoints you hint at, in the context of this Al-Ahli hospital incident?
Well, to people who don't believe in precognition, it sounds like Anthropic had quality control engineers dedicated to their military clients' usage. Basically running through the prompts and inspecting the answers and digging deeper how their chatbots gave those answers. Somebody must have pressed the high-alert button, resulting in Anthropic taking a stance.
Technically the statutes of Rome forbid using human shields.
A nation state bombing US mainland bases sounds rather implausible, although I certainly would prefer that civilian infrastructure to have a minimum distance to military targets, even in the US, even if only to set the right example to the rest of the world.
I do believe there would be value in modernizing the statutes of Rome regarding human shields, which would force nation states to compile machine readable lists of school locations, so that non-existent reported childrens schools and secret childrens schools would be automatically screened.
Keeping the school secret, or reporting a school location too close to a military base would then activate the right of the international community to attack that nation, in order to prevent nation states from using elementary schools etc. as human shields.
IRGC wants nuclear ICBM's. Iran invests heavily in STEM education and physics. The whole population is aware of such goals, the whole population is aware of the adversarial relationship with the Western hemisphere. Imagine your child being allocated the school that was bombed in Iran, but before it was bombed: wouldn't you protest and ask for your child to be allocated to a different school? They risk being the first casualties when the inevitable escalation to war occurs. Clearly in this fun society of Iran, those parents didn't get a choice, and could only pray their kids get through elementary before such a foreign attack occurs.
IMHO, the most damning aspect is that proper, modernized international law clarifying the permitted action-reaction patterns around human shields could have prevented these deaths, by disincentivizing such nations from using kids as human shields.
for small screen sizes and low information density displays, like a watch that updates every second this makes a lot of sense
it would make a lot of sense in situations where the average light generating energy is substantially smaller:
pretend you are a single pixel on a screen (laptop, TV) which emits photons in a large cone of steradians, of which a viewer's pupil makes up a tiny pencil ray; 99.99% of the light just misses an observer's pupils. in this case this technology seems to offer few benefits, since the energy consumed by the link (generating a clock and transmitting data over wires) is dwarfed by the energy consumed in generating all this light (which mostly misses human eye pupils)!
Now consider smart glasses / HUD's; the display designer knows the approximate position of the viewer's eyes. The optical train can be designed so that a significantly larger fraction of generated photons arrive on the retina. Indeed XReal or NReal's line of smart glasses consume about 0.5 W! In such a scenario the links energy consumption becomes a sizable proportion of the energy consumption; hence having a low energy state that still presents content but updates less frequently makes sense.
One would have expected smart glasses to already outcompete smartphones and laptops, just by prolonged battery life, or conversely, splitting the difference in energy saved, one could keep half of the energy saved (doubling battery life) while allocating the other half of the energy for more intensive calculations (GPU, CPU etc.).
I don't believe LED-pixel displays use PWM. I would expect them to use bit planes: for each pixel transform the gamma-compressed intensity to the linear photon-proportional domain. Represent the linear intensity as a binary number. Start with the most significant bit, and all pixels with that bit get a current pulse, then for the next bitplane all the pixels having the 2nd bit set are turned on with half that current for the same duration, each progressive bitplane sending half as much current per pixel. After the least significant bitplane has been lit each pixel location has emitted a total number of photons proportional to what was requested in the linear domain.
What would such predelegated instructions look like, how large is the state space in that flowchart? How effective is control theory with a tiny state space? This doesn't sound like a survival plan, but a self-splintering plan: some military units will capitulate or defect while others fight on, when pushed till the edge, or is there some kind of direct-democracy-within-the-IRGC? that doesn't sound plausible...
Basically sounds like the military from Imperial Japan during the end of WW2, with scattered units continuing to fight, surrender not believed an option, not aware, or in disbelief that Japan has surrendered...
Let's hope it doesn't have to lead to the same conclusion?
The Swedish military famously works the same way (or at least used to) - they're trained to uphold the Swedish constitution themselves regardless of what their leadership says, with the result that they saved many lives in former Yugoslavia despite orders not to intervene: https://www.realcleardefense.com/articles/2017/09/20/trigger... .
This isn't a complicated war. The US can't and won't do occupations, so the only thing you need to do is cause problems till they leave.
Iran doesn't have to conventionally defeat the US military and can't: so they're just not doing that, and instead going after valuable economic targets which are politically sensitive to Americans and impossible to defend since they're risk sensitive.
I can see the following technology replacing motorcycles for communication:
(works up to 20-30km, a bit more if needed)
a) preinstall your fiber optic cable between points A and B (say AA platforms that need/want coordination for distributed passive/multistatic tracking of intruders)
b) when it is torn, send a fiber optic drone from A to B and use its line to replace the torn one (those are flying in Ukraine with bomb payload, now just use its fiber optic reel, you can reuse the drone; not durable, but very cheap and fast repair of radiation-free communication lines)
Today's technology offers so many opportunities ...
Wouldn’t a WiFi mesh network be more reliable in war-torn areas? If you just need communication then actual “internet” is incidental and probably a security risk - just having a fairly secure local mesh network, with nodes covering hot-spot areas, seems like a good idea - it can cross areas where fiber isn’t reliable because of all the war, and it can potentially remove the need for some by-hand communication.
Wifi mesh makes sense in a densely populated area, not over mostly desert.
Also, communication over longer distances (even few km) will add so much latency that it will be unusable for coordinated AA targeting.
Furthermore, all that radiating will just invite bombs from the attacker.
Maybe I was not clear enough about the goal: not "robust command and control communication network", but more of:
quickly and temporarily set up a high-bandwidth low latency communication network to accomplish AA ambush using coordinated mobile passive sensors (a quick radar burst might for initial acquisition might be useful, but probably not necessary).
> And imho, while Kurds accepting to be betrayed by the US for the third time in less than two decade won't have any real long term impact, an independent Baluchistan can easily destabilise Pakistan.
Not to confuse my prediction from prescription, but what prevents all the neighboring (direct or indirect over a sea) nation states from deciding to divide Iran like Germany was during the cold war? Thats not an independent Balochistan, at some point they will want reparations for all the damage, terrorism and intimidation they have incurred from Iran...
At some point the people in Iran will have to be forced to teach their innocent children the equivalent of the Nuremberg trials: there is no excuse in order to stop thinking, just following orders is not a valid legal defense.
Every population has the moral responsibility to keep the local aspiring autocrats in check, because if they don't and external power deconstructs the regime, the onus will be on the population!
Saddam was paid (in chemical weapons, but not only) by the US to invade Iran, it didn't work well for them at the time, despite the MEK helping them with hidden routes and a lot of local support they don't have anymore. The current Iraki leadership isn't stable enough to do the same anyway.
Afghanistan and Pakistan are in a small war that will have some impact on Baluchistan, but official Pakistani ground troops are a no-no, because it will leave ground for the Taliban. Also India invested a lot in Baluchistan biggest port, and Pakistan threatening their investments will probably have them react (India love nothing more than helping Pakistan adversaries). Koweït is too small, Irak Kurds need to secure their autonomous region, and US promised are worth basically nothing. Azerbaijan used Iranian drones and artillery against Armenia like 2 years ago (maybe 3), and Iran apologised publicly after sending a missile to them.
All of this to say: only the US have the manpower and will for a ground invasion.
Probably scale, a few million jews, arabs - qataris and emirates and saudi royalty is unlikely enough to deconstruct Iran, unlike Germany vs multiple comparably or larger sized regional peers.
Iran is 100m large country + 100s millions more shia core / axis of resistance supressed by small regional satraps empowered by outside forces. There are simply 10x more Muslims in region suppressed for decades under same framework where arc of history would would look kindly on Iran+co for destroying US influence and the greater Israeli project and look poorly upon satraps and compradors for failing their spiritual and moral duty of reclaiming the levant. The Nuremburg trials will be reserved for those who failed Islam for secular glitz and kindly on those who protected the faith. Iran simply has the size and spiritual/historic/civilization mandate to win the regional narrative and "moral" war versus gulf monarchs that choose to coexist with Israel. Gulf monarchs who are btw also definitionally autocrats whose contract to bribe populous with petro state proceeds goes away if this war drags on, of all autocrats they are the most likely to fall and least likely to normalize against autocrat regime change. This not to say Iran is "correct/moral" just they have scale and discourse legitimacy Germany didn't.
I don't agree with redirecting towards fossil fuels instead of wind power, but its not really paying TotalEnergies "for not building wind capacity", its more like changing what was ordered on behalf of the population: first the wind power capacity was ordered, then it was stalled and blocked, and now this president and TotalEnergies have agreed to change the order to another type of meal (investing in fossil fuel facilities within the US).
It would seem wise to modify the autoresearch instructions to first estimate the computational costs rigorously and then sort and compare the proposals for human review, and for each actually executed attempt to feed back the computational costs with LoRa adapter?
i.e. perhaps minimal changes to autoresearch can take control for cost-effective research to occur.
Yes but at that point you may as well use a proper hyperparameter tuning framework like optuna if all the LLM agent is supposed to do is do hyperparameter tuning.
Does optuna think abstractly (i.e. use LLM to interpret the code and come up with insights), or just perform hyperparameter tuning experiments on user-indicated parameters?
The latter, but it uses fairly optimized approaches to ensure it selects the best candidates.
If you look at the commits, you can see that all it does is just set different values for different parameters of continuous values: the type of thing that I trust statistics a lot more than reasoning. Optuna can make very informed decisions when making lots of different changes at once, slowly converging towards optimal parameters, where the LLM seems to be throwing stuff at a wall and see what sticks.
What would work best if the LLM would try to approach things on a higher level, ie use Optuna, but reason about better approaches for algorithms and/or data or whatever. But what it ends up doing is tuning parameters manually, only one / a few at a time, extremely inefficient and unlikely to be optimal.
> Yes but at that point you may as well use a proper hyperparameter tuning framework like optuna if all the LLM agent is supposed to do is do hyperparameter tuning.
while the "novelty" of autoresearch is that it may symbolically reason about the computation, analyze the codebase, etc. i.e. a wider search space (harder) but symbolic reasoning.
I am not
> Those of us who paid attention learned to not rush to conclusions, and never, ever trust social media or the western press to overcome or even understand information warfare.
Since you highlight western press can't be trusted to overcome / understand information warfare, would you care to provide some write-ups detailing the viewpoints you hint at, in the context of this Al-Ahli hospital incident?
reply