Don't sell the book, sell your story. People buy stories & how those make them feel as a person, they do not just buy the product.
I have marketed books in the past and they are generally very difficult to sell because discovery is challenging. Cookbook category is especially very competitive and AI tools have made everything worse. So your best bet is to promote your story on social media and try to generate word of mouth. You could try spending some money on Amazon on keywords which are extremely tightly related to your cook book stuff and check if it generates any value, try with a small amount.
What a nice comment!! This has been a big failing of my mental model. I always believed if I was smart enough I should understand things without effort. Still trying to unlearn this....
Unfortunately you must look closely at the details to deeply understand how something works. Even when I already have a decent mental heuristic about how an algorithm works, I get a much richer understanding by calculating the output of an algorithm by hand.
At least for me, I don't really understand something until I can see all of the moving parts and figure out how they work together. Until then, I just see a black box that does surprising things when poked.
It's also important to learn how to "teach yourself".
Understanding transformers will be really hard if you don't understand basic fully connected feedforward networks (multilayer perceptrons). And learning those is a bit challenging if you don't understand a single unit perceptron.
Transformers have the additional challenge of having a bit weird terminology. Keys, queries and values kinda make sense from a traditional information retrieval literature but they're more a metaphor in the attention system. "Attention" and other mentalistic/antrophomorphic terminology can also easily mislead intuitions.
Getting a good "learning path" is usually a teacher's main task, but you can learn to figure those by yourself by trying to find some part of the thing you can get a grasp of.
Most complicated seeming things (especially in tech) aren't really that complicated "to get". You just have to know a lot of stuff that the thing builds on.
It should be useful (assuming the course is good) if you want to get a job programming in Python, although other skills are needed too, but you can build those as you go along.
It can be useful in many ways even if you don't want to get a job programming.
For example, writing simple programs related to any aspect of your personal, professional, business, or social life. You could write small software tools to automate or facilitate tasks in these areas, as many people do.
I get the idea of superconductivity at room temperature in theory. But I don't have enough knowledge to understand in real terms, how will the world change if this is true?
You can't get CT scans very often, because they hit you with large doses of ionizing radiation. Ultrasounds are low-resolution spotlights; you shine them at a particular spot to diagnose something specific.
With this you could get an MRI at your annual checkup. You could diagnose all number of diseases like that, not to mention 95% of cancers. Each year your scan is automatically compared to the previous year, and any sudden changes in morphology can be biopsied. The learning would be revolutionary for medical science as well- right now we have so little data on what kinds of benign growths people have that our best method for figuring out if a mass is a problem is asking if there are any other symptoms. Not to mention entirely new kinds of medical devices would be possible, eg using SQUIDs.
Ground-imaging MRI would also be revolutionized. Archeology, paleontology, geology, mapping resources and finding minerals would experience a quantum leap. You would be able to drive a car through the desert and spot fossils or faults or mineral signatures.
Space travel would become essentially free with the use of launch loops. Which would also make long-distance travel incredibly cheap and practically pollution-free. You would need electricity alone to reach low earth orbit, or to accelerate planes to multiples of the speed of sound.
Grid-level storage, peaker plants and load-following would become nearly obsolete. Superconducting catenaries would connect every nation on earth. Normally plants have to turn off when everyone goes to sleep; now factories in China can be powered by US fission. Canadian homes could be kept warm by Australian solar. HVDC interlinks would be obsolete. We might eventually transition away from AC power entirely.
CPUs could be anywhere from 10% to 50% more efficient. GPUs even more so. Fires, particularly house fires would become less common as wires simply stop conducting when they are overloaded.
> We might eventually transition away from AC power entirely.
This is actually a really good point I hadn't fully considered, but it's right: the primary reason we use high voltage anywhere is because it minimizes resistive losses (and the reason we use AC is because it's easy to transform between voltages).
But most of the stuff in my home doesn't need high voltage - it's all running at 5V or 12V. Or it's a motor which is magnetically driven and depends solely on magnetic field strength (which is independent of voltage).
If all your conductors have zero resistance, then high voltage is obsolete. You could safely run a residential property on 12V power. Home electrical hazards would a thing of the past.
This is drawing completely wrong conclusions from erroneous oversimplifications:
We're using high transmission voltages to keep current down. Superconductors would not change this AT ALL; superconductivity generally breaks down not only with temperature increases but also magnetic field strength (i.e. current).
Switching large currents is also a hassle; especially with non-resistive loads.
And completely changing household electricity architecture is simply not gonna happen just to marginally improve safety, cost/benefit ratio is WAY too high.
A superconductor running at high amperage requiring more superconductor is still a superconductor. The losses you take are zero.
Any amount of cross-section of copper though is not - you take losses at (I^2)*R. You lose power as a square of the current.
There is an enormous difference between using superconductors at high currents and using any normal material.
Obviously the impact of this depends on what the critical current of a hypothetical room-temperature superconductor ends up being...but REBCO tapes achieve current densities of >40,000A/mm2 (at 77K). Depending on what you end up with, the expense and danger of maintaining the high voltage infrastructure could easily be seen as not worth it - particularly if it speeds up the ability to build out and maintain power lines.
Sure, but transission losses are generally a low single digit percentage-- eliminating those will not have much impact, but on the other hand your superconductor is EXTREMELY unlikely to be even close to cost competitive with aluminum/steel core wire.
Even if you could achieve critical currents comparable to conventional high-temperature superconductors at ambient temperature (which appears *highly* doubtful!), keeping high power transmissions lines at human-survivable voltages would be a tremendous waste of super-conducting material.
And even inside homes it seems quite farfetched to me to scale down voltages-- nobody wants to use plugs and switches rated for 200 amps just for their cheap toaster...
Yes, and for short haul that works fine. But for really long haul it doesn't, hence HVDC so that's what you compare with: the situation where it makes a difference such that extra cost incurred doesn't immediately invalidate your option. HVDC is much better comparison material than your average overhead powerline. For the same reason we don't compare bicycles with trucks for long haul cargo but we do compare bicycles with cars for shorter distances and personal travel.
NordLink flows 1400 MW. Wholesale electricity in Germany is roughly $105.
365x24x(1400x.07)x105 = $90 million per year. Adds up to the cost of the total project every 17-22 years. Over 20 years it's $1.8 million per km. If the superconductor is 20 kg/m (2.4" or 6.2 cm width, huge), that's $90 per kilogram.
10x the cost of copper.
It's interesting to see how many assumptions about our world are underpinned by the lack of superconducting material. That also immediately gives you an idea of how transformative (heh.) room temperature superconductors would be.
> But most of the stuff in my home doesn't need high voltage - it's all running at 5V or 12V.
85 volt DC carries the same power as 120 volt AC, but 85 volts DC is essentially safe to touch. The human body has a much lower AC impedance, so it's MUCH more dangerous. DC does still hurt, though.
40-80 volts (see also: split phases) DC is very convenient for most electronics. It's really just things with batteries that want 5-12 volts, but stepping that down isn't too hard.
At the grid scale, it's a question of which is cheaper. If the infrastructure becomes much more expensive (because the wires are SC) then you can save money by using DC (which gives you 41% more power). If its cheaper to use transformers than it is to use more superconductors and semiconductors to convert voltages, they'll do that.
Either way the grid would stay relatively high voltage (10s of kV), because it's just always going to be worth it at that scale to minimize the conductor area.
We use AC because changing voltage levels with it is extremely simple and efficient compared to DC.
In fact the only (practical) way to convert DC voltage levels is to convert to AC, do the level conversion, then convert back to DC.
Believe it or not, DC already is more efficient for energy transfer and why there are already DC high voltage transmission lines. You don't have to deal with reactive parasitics.
But again the killer is that AC voltages are so easy to switch and can by done with >99% efficiency.
Full body MRI scans are only expensive in the West, outside the west you can get one done for $250. This is a labor and regulatory capture problem and not a technology problem and will not be affected in any meaningful way by better superconductors.
Even if it were $250, that's a pretty high cost relative to the annual cost of insurance. It's too high to justify as a routine diagnostic tool. The financial benefit here is that earlier identification would save money in the long term treatment. MRIs don't cure cancer, so the direct benefit only applies to the limited savings on a very small subset of people of the people who actually get cancers that could be identified earlier.
The real benefits are indirect (from the viewpoint of the insurance people who unfortunately pay for it)- quality of life is much better if you catch it earlier, and the medical research benefits are huge.
Realistically, it's also not $250 even outside the US- not for the resolution needed to diagnose cancers. That's below the depreciation cost of a high end (say $1M) machine. 12 scans a day (it takes roughly an hour for an average scan, 12 is per day per machine is pretty average[1]) 7 days a week for 10 years is 43,800 scans. So ignoring interest, labor, and absolutely everything else that's $228 per scan.
A full body MRI takes an hour only for small patients. More realistically 1.5-2 hours.
I know, but my point is that the price doesn’t scale with machine and helium costs, but with labor costs and the level of insurance racket a particular country has.
Obviously it’d be great if we didn’t require helium supply chains to make medical scans, but unfortunately we can’t fix everything with technology alone.
The cost of the machines will drop considerably so it definitely will have an effect. That $250 is still a lot of money for many people and if not covered by insurance and in the developing part of the world it is utterly unaffordable.
> Fires, particularly house fires would become less common as wires simply stop conducting when they are overloaded.
I don't know enough about how this material behaves, but a superconductor "quench"* can be pretty catastrophic. I could see a room temperature superconductor battery causing fires from a quench.
we could do annual scans with current tech. the fundamental limitation of MRI is proton relaxation time, which limits the sampling rate. the path to reducing scan time and thus cost is to use a more sophisticated reconstruction method to reduce the number of required samples. this is being worked on.
i don't have any data here, but I am dubious that a room temperature superconductor will bring down the price of MRI machines. a room temp superconductor only saves you a dewar, about $50k of liquid helium and a cryocooler. you still have to build the rest of the MRI, which is an _extraordinarily_ sensitive instrument
> Fires, particularly house fires would become less common as wires simply stop conducting when they are overloaded.
There is a lot of infrastructure around the MRI in order to support liquid helium storage, cycling, and inter-device pathways. It isn't just what you see in the room.
It might still be a large machine, but a bunch of bottlenecks disappear. With that, it is only a matter of time until a startup develops a much cheaper, smaller, and more efficient device.
Philips is a big player making the assumption that the next generation of MRI will be smaller, cheaper, and more widely available. But, I can confirm smaller players are operating on that assumption as well.
CMIIW but the main thing to make acquisition times more reasonable is higher magnetic field strength. Which, leaving all the technical questions of achieving it aside, comes with other fun constraints like requiring heavy shielding for the room and of course very careful control of what kinds of metallic objects can go near it...
If feasible and indeed not as expensive to produce these materials, high potential for:
- higher efficiency turbines and solar panels - more clean energy for the same investment
- fusion?
- low-energy computing at higher performance, as we learned recently LLMs so far can't take advantage of hitherto zero marginal cost of software anymore
- democratization of advanced quantum computing?
It's all very exciting and in a truly replicable and industrially-feasible scenario I'm starting to feel this could be another 1960s kind of rate of change. One can dream, no? Maybe we can finally get rid of all the doom & gloom stories we tell ourselves and actually do something with these unexpected presents of our times? Think smartness instead of ignorance, (old) Star Trek instead of the latest Fallout fantasy on the horizon? Why not?
These and many more consequential innovations might develop just in time, as climate change is coming at us much faster than we are willing to admit (don't look up).
That said, even with all of that (including fusion) we will still need to cut our co2 emissions; drastically change our lifestyles / minimize consumption and deal with already locked in impacts hitting us sooner than later.
Enthusiastic midnight edit:
Also what's up with graphene based ICs and optical computing advancements? Competition of new old ideas finally come to be realized? What's next? I want a new breed of superconductor enabled Lisp Machines by 2030! Why not home brew "3D print" the whole thing? That should be the ultimate target here! The handling of "open source" lead would probably suck though %D.
I guess Alan Kay wouldn't be enthused by such a Lisp Machine renaissance in principle yet still stand with his "the best way to predict the future is to invent it" credo.
Let's predict a future for a planet that shifts back into balance!
You know all those bird scooters people leave lying around cluttering up the sidewalk? With a room temperature superconductor they will become hover scooters cluttering up the sidewalk!
And all of the high voltage transmission lines we want to build but can’t because of permitting reasons would have zero energy loss if we actually built them, which we won’t.
Yes and no. There's no possible way to have a superconductor be an actual hoverboard out in the broader world - the magnetic field of the earth is just nowhere near strong enough. You can make specialized areas where it would work, though. It's even been done already with LN2 cooled superconductors - https://www.theverge.com/2015/8/4/9091951/lexus-hoverboard-v...
SC could help enable nearly lossless transmission over power in HVDC lines, but HVDC lines are already significantly more efficient than our regular ones and we don't build them for a variety of reasons, so it might not make much of an impact there for regulatory/NIMBY/etc. type reasons.
We would first need to build a superconducting track for the hover scooters (containing magnets) to ride on. Assuming we ever figure out how to mass produce the superconductor inexpensively.
Perhaps if the track were arranged in a grid-like pattern, the scooter could use superconducting electromagnets to accelerate and steer.
I would actually welcome scooters that only work on specialized tracks. Walking safely on the sidewalk will become as easy as staying away from those tracks.
I don't think anybody can really say for now, but if the price of it comes down enough, there are two rough categories of things that can happen:
* devices that currently use superconductors don't have to use cooling anymore, and so become much cheaper to build and operate (MRI machines, certain sensors, high-power magnets for things like fusion research, big generators, big motors). This is a pretty solid bet.
* devices where superconductors would be an improvement, but currently don't make economic or practical sense. These are almost certain to crop up, but which ones will pan out is IMHO very speculative.
In the latter category, things like computing chips, more sensors, certain art works (sculptures with permanently levitating parts, how awesome!), smaller motors and generators seem plausible.
But there is likely whole categories of things we haven't thought of that could benefit from either zero resistance or rejecting magnetic fields.
It means we won't lose MRIs when the earth finally runs out of helium as they require liquid helium to run the superconductors that generate the magnetic field. We're running out of helium with no way to replenish it. Helium is the only element on the periodic table which is a non-renewable resource on Earth.
So MRIs will get much cheaper, and they could end up being as cheap as taking an x-ray today.
Based on our current data LK99 isn't a replacement for helium cooled superconductors for the same reason YBCO and similar aren't - critical current/critical magnetic field aren't in the right ranges for what we need out of MRI machines.
I do think it's too early to say one way or the other what all of this ends up looking like, so we might find that purer/larger samples have better properties than what was measured so far, or the discovery puts us on the trail of other RTAPS in the same class that might be better for these purposes.
I never claimed LK99 had the properties needed for MRIs, the question I was answering was "how will the world change with room temperature superconductors", and my answer is valid in the context of the question.
If the earth runs out of helium we will just use low field MRIs that don't require superconductors (see eg https://www.nature.com/articles/s41467-021-25441-6 ). Their resolution is lower than that of high field MRIs, but they still seem to be a useful diagnostic tool.
I suppose you could count elements like gold, too. Since we technically can produce extremely radioactive gold in small amounts but the cost is so prohibitive it would never make sense to.
In a thousand years people are gonna look back at us idiots filling balloons with helium and letting them disperse into the upper atmosphere and shake their heads at how stupid we were.
I think you mean lossless transmission? If we can actually replace transmission lines effectively with it, unclear whether it's practical for that yet.
I'm thinking coil guns instead. Room temperature superconductors don't solve the rail erosion issue with railguns, but I think should greatly increase the performance of coil guns.
My thinking is that zero resistance through the projectile itself and through the rails would help, but you still need to make an electrical connection between the projectile and the rails. Either this is done with a plasma arc or physical contact, but either of these causes erosion of the rails even if there is no electrical resistance through the rails or projectiles. Am I missing something?
Coil guns seem like they'd last longer, wonder what the performance figures for coil vs rail are.
For future rail guns, they'd just have replacement rails available as they do barrels for tanks/artillery guns (that wear out after about 1000 shots afaik).
if he means that, I do not see it happening so easily,
Energy storage in a superconductor is done in the form of magnetic field, a superconducting induction coil (SMES), whose density of energy storage per kilogram is highly inferior to a capacitor and a supercapacitor, which stores energy in the form electric fields, and whose density of energy storage per Kg in turn is very inferior to chemical batteries.
The magnetic fields are charged and released quicker in inductors than in capacitors (and than in chemical batteries), also the material have a longer life, and the rate of self-discharge is sightly inferior in superconducting inductors, nevertheless the density per kilogram -and to administrate such sudden energy release- limits very much the applications.
If LK99 becomes true, and is improved much (as the electrical current in the paper is limited to milliamperes range), at middle term I don't think on it happening.
If at future is achieved superconducting through nanowires, with inferior weight to batteries, may be, but in a car for example would need the added weight of metallic "magnetic shields" for health security.
IMHO, I don't see it beyond stationary applications.
This is consistent with my understanding as well. I don’t see instant battery charge being something that this enables. Current battery charge rates are limited by battery structure degrading more rapidly at high charge/discharge rates, so this wouldn’t really have any direct effect.
Would you mind sharing, what level of programming & mathematical background do I need? I know basic python (read python for data analysis) & currently half way through elements of statistical learning. What else do I need to learn?
> half way through elements of statistical learning
If you're able to make it that far on ESL, mathematical background certainly won't hold you back when learning anything "neural networks" related. Specially not a spelled-out practical intro.
> I know basic python (read python for data analysis)
You may want to get more comfortable with programming in general (outside of the data analysis realm), but you can learn everything you're missing while watching Karpathy's series (and referencing the python docs).
I totally agree with you. For this reason, I don't follow mainstream outlets about climate change. The most balanced source I have found so far is this substack by Hannah Ritchie. Here is one of those balanced article - https://www.sustainabilitybynumbers.com/p/15c-warming-tempor...
Yeah, just saying suffering is bad without any context is not helpful. Apart from being unavoidable, certain kinds of suffering produces happiness for people down the road. For example, running & training for a marathon was suffering but the reward and experience of doing it was life changing, taught me so much about myself.
I feel sad for the people of Myanmar. I feel sad for Aung San Suu Kyi and all freedom fighters who have sacrificed their whole life and family to better Myanmar.
My personal experience is Myanmar is a wonderful country full of a wonderful people, and vibrant culture, despite living in the shadow of ethnic violence and a brutal military regime.
It’s hard to say how this will play out. For many here it’s a mystery why military did this, as they already had power over state business, 25% military appointed legislature, control of courts, etc. and the benefit of hiding behind fake democracy to get foreign investment. So people are confused as to what their motive is, as they already had the best of both worlds.
> So people are confused as to what their motive is
Economic prosperity for the populace is dangerous for a dictatorship. Myanmar is still a poor country which exports mainly things dug out of ground. If the situation changes and now the military is financed by businesses and tax-payer money, they'll have to answer to those.
This is bad news for the military leaders and their gang. They'd much better have a small income from these resources and a poor population than being ousted. (and probably face prosecution)
>While failures to reform the country can be considered by many to be at the disadvantage of NLD, in a more cynical reading, it might actually be to the party’s advantage. Ironically, it helps to convince the people that there are still influential dark forces (read: the military) that deter the party from reforming the country. [1]
Do you believe that your consideration of the previous military power is widely understood and that the above quote explains the success of the NLD in the election?
Do you feel San Suu Kyi has been prevented from acting to protect ethnic minorities due to the influence of the military?
If so, is there reason to believe the military would perceive the beyond landslide victory in favor of the NLP as Likely to disrupt or erode their existing position of power?
Most international press is dangerously misinformed about structure of government in Burma. For example, constitution (created by military) gives military control of 25% legislature seats, president, and courts. The military doesn’t answer to any civilian authority. Civilian government can’t reform constitution or appoint critical positions without military approval.
Almost everyone in Burma hates military and understands that civilian government can’t do everything they want.
Yet all the international coverage seems ignorant of these facts, condemning ASSK for things she doesn’t control, like military operations in Rhakine against Rohingya. They don’t understand that the reason she defended military at UN is only outcome of not defending them would be sanctions which just further harms civilian government and people. She had no choice.
People here in Burma are still confused as to motivation for coup, since military already had best of both worlds: ultimate power of law and courts, and no sanctions because of farcical democracy. Why they would risk sanctions to do this is a mystery.
> Why they would risk sanctions to do this is a mystery.
Sorry for offering wild conjecture, but the first thing that comes to mind is some kind of a payout to generals from China, who is hoping for sanctions from the rest of the world so that they can be the sole influencer in the country.
Or some variation on that theme: influence from China.
Or perhaps just some power hungry generals longing for the old days <shrug>
No so crazy. Many believe Chinese interests like behind what’s been happening in Rakhine since 2017 as well. The rest of the world shunning Burma benefits the CCP more than anyone other than the Myanmar generals.
> For many here it’s a mystery why military did this
Big Tech started banning Myanmar government officials from their services that they didn’t approve of. Now they cry about suspension of democracy? The claims of a rigged elections in Myranmar sounds eerily familiar...
I have marketed books in the past and they are generally very difficult to sell because discovery is challenging. Cookbook category is especially very competitive and AI tools have made everything worse. So your best bet is to promote your story on social media and try to generate word of mouth. You could try spending some money on Amazon on keywords which are extremely tightly related to your cook book stuff and check if it generates any value, try with a small amount.