All of micpie's Comments + Replies

Interesting article, especially because I’m currently rereading some decision making material in the light of some LLM projects.

I think a very interesting part of your discussed setups is how the world model is defined in detail.

I see it as something that is already involved in our perception with some major restrictions, i.e., incomplete observability and some kind of biological hard-coded guidance system (“feelings”). This view still allows for a factored model, but it comes with different dependencies between the building blocks, i.e., between the value... (read more)

Very interesting article! A big part of the outlined techniques I also kind of discovered by trial and error.

I recently also started to work more on my math skills and putting in the mathjax was getting very time consuming. To reduce that bottleneck I can recommend this small app:

PS: I stumbled once over this collection of proofs without words and they make great Anki cards to build up intuition:

thank you! This is amazing. I wish I'd known about this sooner.

A book that goes very much in that direction with small but impactful chapters is:

“Chop Wood Carry Water”

I really enjoyed (and as I’m now reminded of it, I will have a look at it again) and I guess you could like it too.

(I found it via and the very good review got me interested in it.)

Out of curiosity I had to look at the FT chart with relative numbers (= seven-day rolling average of new cases (per million), by number of days since 0.1 average daily cases (per million) first recorded):

Linear scale: 

Log scale: (read more)

I guess a deficit in quantitative reasoning is just one of the contributing factors.

Another contributing part, I keep thinking about a lot, is the role of social media during the pandemic. Social media is making money by engaging people. The longer people are on your platform, the more data you can harvest and the more advertisements you can show them, resulting in more revenues. And the more data you have, the better you can target the ads, and so on. The best way to drive up engagement is to promote controversial posts (the more extreme the better, you l... (read more)

To grow the virus and inactivate them by chemical reagents is the standard setup for (older) influenza vaccine systems that are egg- or cell-culture-based. From such a setup whole-virus- or subunit vaccines can be derived.

A nice graphical guide on COVID-19 vaccines:
Here's a snippet from it... It also links to WHO's spreadsheet of COVID-19 candidate vaccines, which looks like a good resource in its own right (although this particular copy is probably a bit old).
Answer by micpieApr 01, 202021

To produce a vaccine, you will need at least:

  • Vaccine (or at least a rough idea what your vaccine will be going to look like)
  • Large scale production process (which is not the same as a bench scale process)
  • Raw materials
  • Facility (see below)
  • Staff (see below)
  • A working quality system (including an ok from the regulatory agencies)

In biopharmaceutical production you have two kind of extremes in the facility design:

  • "Stainless steel" facility: This is usually the option if you are going to produce the same product forever. It needs a higher upfront investme
... (read more)

No direct prediction from my side but a link to a report:

The full PDF report (linked on the website) has on page 15 a overview of possible outcomes that could be a basis for discussion.

Interesting comment on a (maybe) new symptom, i.e., loss of smell and taste for several days, of a COVID-19 infection in an interview of a MD with focus on Virology in Germany:

Google translation of the interesting part:

"Almost all infected people we interviewed, and this applies to a good two thirds, described a loss of smell and taste lasting several days. It goes so far that a mother could not smell the full diaper of her child. Others could no longer smell their shampoo, and food began to taste bland. We cannot yet tell exactly when these symptoms ... (read more)

What I know from clean rooms in the biopharmaceutical production is that you avoid there cardboard at all because there is no straightforward way for disinfection (besides the particulate contamination that comes with them). Therefore, one approach is to remove the cardboard as soon as possible and put it away (and wash your hands afterwards).

Edit - Additional comment to make the statement more precise:

There is no straightforward way for disinfection of cardboard without destroying it, i.e., the cardboard soaks in the cleaning agent and will disintegrate.

[This comment is no longer endorsed by its author]Reply

Thank you very much for the perceptron and the Explorable Explanations link. I didn’t know the EE site. If you can recommend similar material, I would be very interested in it! :-)

Michael Nielsen also has some great stuff. Especially his and neural networks one.