@ckfinite.bsky.social
If that's how ads work, then it makes it impossible to tell whether the answer was rooted in an ad, in the result from a tool call (e.g. search), or in the training data.
December 21, 2025 at 10:29 PM
If that's how ads work, then it makes it impossible to tell whether the answer was rooted in an ad, in the result from a tool call (e.g. search), or in the training data.
What I think is going to be very interesting is how the LLM providers deploy ads. The most obvious way to do it (injecting ad prompts through a RAG into the LLM's context) is also devastating to the utility of the LLM in the first place.
December 21, 2025 at 10:28 PM
What I think is going to be very interesting is how the LLM providers deploy ads. The most obvious way to do it (injecting ad prompts through a RAG into the LLM's context) is also devastating to the utility of the LLM in the first place.
There's a weird assumption that a reverse transcriptase comes from... somewhere... that's never specified. I don't really get it as an argument.
October 27, 2025 at 8:28 AM
There's a weird assumption that a reverse transcriptase comes from... somewhere... that's never specified. I don't really get it as an argument.
Another issue I'd point out that I like to keep harping on is that launch is already much cheaper than almost any payloads are. Unless you're building a megaconstellation (or a megasatellite, which are as yet unflown) building your satellite is going to be the big capital sink now, not launch.
January 3, 2025 at 6:10 AM
Another issue I'd point out that I like to keep harping on is that launch is already much cheaper than almost any payloads are. Unless you're building a megaconstellation (or a megasatellite, which are as yet unflown) building your satellite is going to be the big capital sink now, not launch.
To be honest, I've found what I learned through my education extremely useful and I use it every day. In hindsight, I kinda wish I could have done more (in particular, I'd love to have studied control systems, numerical linear algebra, and EM field propagation).
December 13, 2024 at 8:28 PM
To be honest, I've found what I learned through my education extremely useful and I use it every day. In hindsight, I kinda wish I could have done more (in particular, I'd love to have studied control systems, numerical linear algebra, and EM field propagation).
A nigh-on identical process to UHT milk is canning or jarring, which similarly relies on initial sterilization followed by hermetic packaging for longevity. Normal, non-UHT pasteurization is simply done at lower temperature and is not as comprehensive as UHT/jarring/canning are.
November 29, 2024 at 9:10 AM
A nigh-on identical process to UHT milk is canning or jarring, which similarly relies on initial sterilization followed by hermetic packaging for longevity. Normal, non-UHT pasteurization is simply done at lower temperature and is not as comprehensive as UHT/jarring/canning are.
If the UHT process fundamentally changed the milk to be inimical to life so as to enable shelf stability then it would not need to rely on the integrity of the packaging for said stability. Instead, it kills all the bacteria that were originally there and then the packaging keeps new ones out.
November 29, 2024 at 9:07 AM
If the UHT process fundamentally changed the milk to be inimical to life so as to enable shelf stability then it would not need to rely on the integrity of the packaging for said stability. Instead, it kills all the bacteria that were originally there and then the packaging keeps new ones out.
I think that the causality is backwards here. It's the bacteria eating the milk that causes it to go bad, rather than vice versa. As an easy example, UHT milk is fully shelf stable so long as the packaging is intact but open it and leave it out and it'll be bad within the day.
November 29, 2024 at 9:05 AM
I think that the causality is backwards here. It's the bacteria eating the milk that causes it to go bad, rather than vice versa. As an easy example, UHT milk is fully shelf stable so long as the packaging is intact but open it and leave it out and it'll be bad within the day.
The issue is ultimately that the milking process is a fleshy, organic, one and all such processes have substantial risk of contamination. It can be mitigated, but the extent to which it can be controlled is insufficient to meet modern standards for risk from foodborne illness.
November 27, 2024 at 8:52 AM
The issue is ultimately that the milking process is a fleshy, organic, one and all such processes have substantial risk of contamination. It can be mitigated, but the extent to which it can be controlled is insufficient to meet modern standards for risk from foodborne illness.
Just... generally speaking, commercial refrigeration conditions are sufficient to keep the microbial growth in raw milk (and pasteurized milk) low, so the CFU load in raw milk off the store shelf is not much bigger than that what came out of the milker.
November 27, 2024 at 8:51 AM
Just... generally speaking, commercial refrigeration conditions are sufficient to keep the microbial growth in raw milk (and pasteurized milk) low, so the CFU load in raw milk off the store shelf is not much bigger than that what came out of the milker.
What work has been done on raw milk is primarily in the pre-pasteurization phase (since it needs to be stored and transported for pasteurization). For example, vuir.vu.edu.au/33830/1/vith.... That data suggests that the cold chain has a similar effect on raw milk to pasteurized milk.
vuir.vu.edu.au
November 27, 2024 at 8:49 AM
What work has been done on raw milk is primarily in the pre-pasteurization phase (since it needs to be stored and transported for pasteurization). For example, vuir.vu.edu.au/33830/1/vith.... That data suggests that the cold chain has a similar effect on raw milk to pasteurized milk.
As far as I'm aware, there hasn't been a systematic analysis done of the cold chain for raw milk and it bounds the answers anyways; if the extra bacteria in raw milk grow slower under cold chain conditions then it will perform no worse than pasteurized milk and if they grow faster then, well.
November 27, 2024 at 8:42 AM
As far as I'm aware, there hasn't been a systematic analysis done of the cold chain for raw milk and it bounds the answers anyways; if the extra bacteria in raw milk grow slower under cold chain conditions then it will perform no worse than pasteurized milk and if they grow faster then, well.
The problem is that guaranteeing lack of fecal contamination at scale is practically impossible (or similarly that no cow is infected). Instead, a stochastic analysis is needed to consider how probable it is that the inevitable contamination will make it through to cause illness.
November 27, 2024 at 8:37 AM
The problem is that guaranteeing lack of fecal contamination at scale is practically impossible (or similarly that no cow is infected). Instead, a stochastic analysis is needed to consider how probable it is that the inevitable contamination will make it through to cause illness.
Furthermore, the length of the cold chain (that is, the time from when the milk leaves the cow to when it's on store shelves is not particularly long). journals.asm.org/doi/pdf/10.1..., for example, reports that absolute maximum is 5 days and that most microbial growth happens in domestic settings.
journals.asm.org
November 27, 2024 at 8:34 AM
Furthermore, the length of the cold chain (that is, the time from when the milk leaves the cow to when it's on store shelves is not particularly long). journals.asm.org/doi/pdf/10.1..., for example, reports that absolute maximum is 5 days and that most microbial growth happens in domestic settings.
(sorry, linked the wrong paper in the original reply).
This is a particularly interesting outbreak from this perspective because the virus cannot replicate in the milk (there are no cells for it to infect there). Instead, the virus "lives" dormant in the milk until it's consumed.
This is a particularly interesting outbreak from this perspective because the virus cannot replicate in the milk (there are no cells for it to infect there). Instead, the virus "lives" dormant in the milk until it's consumed.
November 27, 2024 at 8:21 AM
(sorry, linked the wrong paper in the original reply).
This is a particularly interesting outbreak from this perspective because the virus cannot replicate in the milk (there are no cells for it to infect there). Instead, the virus "lives" dormant in the milk until it's consumed.
This is a particularly interesting outbreak from this perspective because the virus cannot replicate in the milk (there are no cells for it to infect there). Instead, the virus "lives" dormant in the milk until it's consumed.
Milk can absolutely be contaminated to unsafe levels straight out of the cow. Fecal contamination of the milk is the most common problem, though this is direct expression of the virus by the epithelial cells that line the avioli in the mammary gland (www.nature.com/articles/s41... see fig. 1 (a)).
www.nature.com
November 27, 2024 at 8:18 AM
Milk can absolutely be contaminated to unsafe levels straight out of the cow. Fecal contamination of the milk is the most common problem, though this is direct expression of the virus by the epithelial cells that line the avioli in the mammary gland (www.nature.com/articles/s41... see fig. 1 (a)).
The main danger of milk is in microbiological contamination, as in this case. This is comparatively easy to control in industrial settings, with a combination of cleaning, good mechanical design and materials selection, and refrigeration. The primary source of contamination is the cow side.
November 27, 2024 at 6:07 AM
The main danger of milk is in microbiological contamination, as in this case. This is comparatively easy to control in industrial settings, with a combination of cleaning, good mechanical design and materials selection, and refrigeration. The primary source of contamination is the cow side.
Assuming something hasn't gone horribly wrong - such as a failure of the refrigeration plant, for example - there isn't a lot in the processing of milk that introduces danger to consumers. The milk production process is very simple: it's pasteurized, the cream is removed, and then homogenized.
November 27, 2024 at 6:05 AM
Assuming something hasn't gone horribly wrong - such as a failure of the refrigeration plant, for example - there isn't a lot in the processing of milk that introduces danger to consumers. The milk production process is very simple: it's pasteurized, the cream is removed, and then homogenized.
No; it's entirely a genetic thing where you have or don't have a specific mutation. Everyone is born making lactase, but depending on your specific genetics that gene stops being expressed as you get older. Some populations are more likely than others to have the mutation that keeps it working.
November 27, 2024 at 6:03 AM
No; it's entirely a genetic thing where you have or don't have a specific mutation. Everyone is born making lactase, but depending on your specific genetics that gene stops being expressed as you get older. Some populations are more likely than others to have the mutation that keeps it working.
In my opinion the main practical utility of raw milk is in the manufacture of hard (aged) cheeses, which accomplishes the same aim as pasteurization. However, I find the qualitative difference between raw and pasteurized cheese limited compared to other cheese manufacturing process variables.
November 27, 2024 at 5:59 AM
In my opinion the main practical utility of raw milk is in the manufacture of hard (aged) cheeses, which accomplishes the same aim as pasteurization. However, I find the qualitative difference between raw and pasteurized cheese limited compared to other cheese manufacturing process variables.
Pasteurization doesn't substantially affect lactase content in milk, mostly because there isn't any. Lactase in milk would facilitate its enzymatic breakdown before it could be consumed, which would somewhat spoil the point. The source of lactase is one's own native production of the enzyme.
November 27, 2024 at 5:55 AM
Pasteurization doesn't substantially affect lactase content in milk, mostly because there isn't any. Lactase in milk would facilitate its enzymatic breakdown before it could be consumed, which would somewhat spoil the point. The source of lactase is one's own native production of the enzyme.