I also don't believe in the 'bitter lesson' when extrapolated to apply to all 'AI application layer implementations' - at least in the context of asserting that the universe of problem scopes are affected by it.
I think it is true in an AI research context, but an unstated assumption is that you have complete data, E2E training, and the particular evaluated solution is not real-world unbounded.
It assumes infinite data, and it assumes the ability to falsify the resulting model output. Most valuable, 'real world' applications of AI when trying to implement in practice have an issue with one or both of those. So in other words: where a fully unsupervised AI pathway is viable due to the structure of the problem, absolutely.
I'm not convinced in the universality of this. Doesn't mean the core point of this essay on the futility of startups basing their business around one of the off the shelf LLMs isn't valid - I think for many they risk being generalized away.
Li-ion is a very big umbrella of battery chemistries, and within can have broadly different safety and stability profiles.
More important is that Li-ion cells are no joke, and are not equivalent to ‘AA’ or ‘AAA’ battery formats. You can NCM, NCA, LFP, Cobalt chemistries and blends within - not even accounting for differing mAH and charge/discharge capacities. 18650 is simply a form factor.
Better labeling, specing, regulated and certified repair processes are just a start for solutions as it is currently a Wild West.
However, statements like ‘Li-ion should be replaced ASAP’ do not factor in that… lithium ion as a platform is not going anywhere and has essentially won at least the next half century for 18650 type form factors and applications. Moving away is not a reasonable solution.
Sorry but that just isn't right. The battery you listed is an just an NMC battery , the listing pages are just lazy and don't list it. There isn't just a plain lithium ion cell. Also lithium polymer just means that a polymer electrolyte is used instead of a liquid electrolyte.
That's pedantic. They're sold that way by the truck load and anybody that buys them for DIY purposes knows that this is how they are designated. People are buying based on the working voltage nominally 3.6V, Lithium Ion chemistry, battery capacity, form factor. Not on the internal details of the battery, which are not normally listed in sales documentation on consumer oriented outlets anyway.
Yes, you can go into the details of the exact chemistry, the packaging, the manufacturers specs and so on but for bulk DIY purposes this is the reality. The detailed chemistry obviously has some implications for safety but because you will have no idea what is in a pack unless you open it (and plenty of times the cells are unmarked) you just have to assume the worst. That's the safest.
> People are buying based on the working voltage nominally 3.6V, Lithium Ion chemistry, battery capacity, form factor. Not on the internal details of the battery, which are not normally listed in sales documentation on consumer oriented outlets anyway.
Lithium Ion chemistry is the internal details of the battery. Which denotes the type of lithium ion battery, there is no generic lithium ion battery. It will fall under a certain category (Ex LFP, NMC, NCA).
Sure you can go buy unmarked cells on ebay but any reputable website should list the battery chemistry on their website or have a link to the datasheet. It is no different than buying any other electronic component.
I've refurbished many packs and built my own, for cameras, e-bikes and laptops. I've yet to find anybody other than myself to be interested in the exact chemistry of the batteries involved. What they're interested in at the highest level is whether it works, and second to that how much it will cost them. I'd be happy to bet that most people that use these batteries would not be able to figure out what exact composition a cell has nor would they care.
It is very clear to me that there is a vast gulf between DIY and professional application, and for the latter I'm 100% sure that everybody there would be able (and should be able) to make the distinction. The number of hobbyists that can do so is going to be very low.
For maximum cringe, have a look at this youtube video and spot the potential problems:
You are missing the core technology point that drives the underlying thesis here - ‘every manufactured good, gets cheaper per unit as you make more and more of it’.
Whether this is from the lab, a recipe of plant inputs, or some other currently unknown thing - if you are able to develop and commercialize a product around a ‘specific technology’, you can drive the cost of that down through experience and scale.
Animals are a VERY thermodynamically inefficient way of converting plants to ‘high quality proteins’. We have automated the crap out of the food processing system. There are no material efficiencies left to be had. Compare that to options which are on paper thermodynamically superior (I.e. not supporting the life of an animal to only use their muscle tissue).
By taking a technology with a much higher theoretical efficiency, and then scaling that up - creates the classic technology disruption scenario. ->‘It’s the cheaper version of X commodity, why wouldn’t I choose that?’
> Animals are a VERY thermodynamically inefficient way of converting plants to ‘high quality proteins’.
That would be pertinent if they were eating the same foods, but they don't eat the same foods. They eat the stuff we won't eat, produced largely either as a byproduct of the production of the foods we do eat or produced on lands that cannot support the foods we eat.
The more efficient we get in producing plants to eat, the more efficient byproducts there are for animals to eat, so you get a continuous relationship of meat becoming cheaper as other foods become cheaper. Economies of scale can only get you so far when the raw material inputs are your main cost centre.
The industry will have to move towards using those 'waste' products in order to be competitive, but at that point you're essentially just replicating animal processes and you're up against a 'machine' doing the same that has had millions of years to develop itself.
> Animals are a VERY thermodynamically inefficient way of converting plants to ‘high quality proteins’. We have automated the crap out of the food processing system. There are no material efficiencies left to be had. Compare that to options which are on paper thermodynamically superior (I.e. not supporting the life of an animal to only use their muscle tissue).
This really is such a cool take on it that it bums me out to say…
>why wouldn’t I choose that?
…because I just[0] don’t wanna eat fake meat.
[0] On a very deep, perhaps primal level. I can feel it rooted so far in my brain it’s hard to convey. Rather eat peanut butter at every meal.
That is because CdTe is actually a stable molecule that isn’t water soluble, turns out the problem of panels from a toxic leeching perspective is largely from the small amounts of lead used at soldering points which is a common point with most silicon PV panels.
No reason tin can’t be used instead/lead free panels can’t be made except for the saving of a few pennies by skimping on the solder - no regulations to ensure they should be lead free. Unfortunately waste processing decades into the future isn’t often accounted for and those few pennies at design time add up across hundreds of thousands of panels on a single site…
Also, typical lead-free solders oxidize in air at their soldering temperature much more readily than near-eutectic Pb-Sn at its respective soldering temp.
That oxide film tends to interfere pretty badly with wetting of surfaces being soldered. In other words - it's much easier to end up with a "dry" joint - even with adequately increased temperature - unless better and/or more flux is used.
Not doubting the facts you've mentioned, but in my experience I've had no problems getting the surfaces to wet, though that may just be due to using rosin core solder that happens to have a good flux in it.
Not to be incredibly nit-picky, but the costs of the product right now don’t matter nearly as much as the ability to scale production and variable unit costs.
As a ‘hard tech’, they need to work themselves through the technology readiness framework before the cost of a particular material even tells you anything of value.
The competition would be carbon neutral pesticide free soybeans, which I'm not sure are readily available? Soybeans also typically use energy intensive nitrogen fertilizers, which I'm not sure this process requires. (It might, not sure.)
Also, soybeans need to be shipped from fertile areas. This could be produced in situ in infertile areas.
I'm not saying it's a panacea, but it has some benefits over traditional intensive agriculture.
Is it actually standard practice to use nitrogen fertilizers with soybeans? Soy is a nitrogen fixer and I've read that nitrogen fertilizers often reduce yields for soybeans because it interferes with nodulation and undermines that plant's nitrogen fixing capacities.
Apparently soybeans are often grown in a double-cropping rotation, in which maize is planted first (along with N and P fertilization), and then no additional fertilizer is applied to the following soybean crop.
This seems to be a comprehensive discussion (for South Dakota farmers anyway, pdf):
Making use of nitrogen fixing plants alleviates the need of fertilization. An easy one: clover. Bees also love clover. However you can explore other systems of cover cropping (I recall Diakon radishes draw up nitrogen from down low, up to the surface) which you terminate before seeding your main crop.
> The competition would be carbon neutral pesticide free soybeans, which I'm not sure are readily available? Soybeans also typically use energy intensive nitrogen fertilizers, which I'm not sure this process requires. (It might, not sure.)
No it's not. This is the kind of thinking that so many new founders screw up with.
The competition is what users are willing to buy instead of you. That's regular soy beans, not whatever market you decide it should be. Don't think this way.
I'm thinking like a consumer, specifically me as a consumer. I do not know how many consumers are like me, but I'm always looking for vegan, well rounded, non industrially farmed, pesticide free protein sources. So far, the leading candidate is pea protein, and while I love soybeans and tofu, it's hard to find soy that is ethically sourced. (Not encumbered with awful genetic patents, for instance.)
The only thing I think that matters is the cost. To sell at scale and displace other foods, it is going to have to be as cheap as chicken. There is no point building a large factory that can produce this stuff at quantity at the price of wagu beef, because there is no market for that product. Even to vegans you are competing with tofu. Cost is why Quorn remains a niche product.
‘Sustainable’ in the food context is much more around total resource use (land, water, materials), waste products (ghgs and waste water) rather than a primary focus of energy efficiency.
Besides - heatpumps are largely the exception in thermodynamics, expecting anything more than 100% efficiency for most non-heat moving applications should not be expected.
Capturing CO2 is not ‘energy cheap’ and therefore it is best when and where possible much of what we start capturing to rid it from the carbon cycle rather than re-introduce it.
a) it will not necessarily last longer, li-ion in the right conditions (particularly more stable chemistries) have superior cycle life experience in practice
b) yes, newest batteries decay slowly over time and don’t ‘just die’ - but this is the same as li-ion
c) safety is mitigated in the modern setting by utilizing the same charge controllers used to keep EV car batteries safe. Remember most North American homes with a natural gas furnace literally is a controlled blast of explosive gas without a smell that people just have and don’t think about - as the tech goes on the learning curve the management issues (I.e. disconnecting faulty cels individually) seems like a plausible end solution, even if that is necessary.
Lastly, while NiMH does not use soon to be very scare processed lithium, NiMH uses very large amounts of nickel, cobalt and comparatively exotic ‘rare earth metals’ who’s production should be expected to be even more challenging to scale.
All that said - I am all for us scaling up all of these technologies so that each can fit within a particular market niche based on inherent pros and cons.
Hypothetically this is not a binary option, they could significantly ramp up the strictness of 3rd party sellers increasing the costs of making multiple, temporary fake seller accounts more challenging.
That said though - this would clearly impact profits proportional to the level of strictness/barriers.
For getting an intuition on it, I would recommend the clip from HBO’s Chornobyl which ELI5’s a nuclear meltdown. [1]
If the reactor is air cooled, that can be translated as “the heat can just burn off” as there isn’t a liquid cooling system, reducing this failure mode.
As mentioned in the video, reactions are about balance. If a ‘safe’ SMR gets shut off or has its systems fail, you would want the accelerators / enablers of the reaction to disappear, and roll out the reaction.
This is unlike most fission reactors in use today which are a ‘balanced dance’ where lots of the design is focused on reducing the speed of the burn (like a constant stream of water on a fire) rather than say controlling the oxygen of a fire with limited fuel. SMRs aim be more of the “control the oxygen” type of reaction rather than “keep the fire cool” kind of reaction. (Forgive my vast oversimplification, but this is my intuition on it).
I think it is true in an AI research context, but an unstated assumption is that you have complete data, E2E training, and the particular evaluated solution is not real-world unbounded.
It assumes infinite data, and it assumes the ability to falsify the resulting model output. Most valuable, 'real world' applications of AI when trying to implement in practice have an issue with one or both of those. So in other words: where a fully unsupervised AI pathway is viable due to the structure of the problem, absolutely.
I'm not convinced in the universality of this. Doesn't mean the core point of this essay on the futility of startups basing their business around one of the off the shelf LLMs isn't valid - I think for many they risk being generalized away.
reply