Hi, I work with generative machines. Everything from Markov chain generators to GPT-3. I’ve trained and tuned many models with GPT2 and 3, all with the intent of simulating human interaction.
I know a fair bit about generative machines, both how they work, and how to tune and interact with them to get particular results.
I need you to hear this: they do not know or understand anything. They are complex probability tables.
1/
This build doesn't filter as much air as I'd like.
On the table you can see a Silverstone AP182 fan floating a ballon, drawing air through a Levoit Core 400s filter. That's about 2.5x more powerful than what's in the lamp. I plan to use that instead for the next iteration.
Behold "Glaze", a software tool that subtly alters the pixels in a piece of art online 🎨 ...
... in a way that isn't detectable by the human eye 👀 ...
... but renders the image useless for training a visual AI. The AI can't detect the artists' style, and thus can't mimic it and copy it. 🤖 ✖️
It's item #1 in my weekly "Linkfest" newsletter here 📫 : https://buttondown.email/clivethompson/archive/linkfest-6-gps-on-the-moon-8-bit-jane-austen-and/
(Subscribe here! https://buttondown.email/clivethompson)
As #norovirus ramps up in the US, a couple reminders I personally needed:
- Norovirus is an *asshole*. It can live (and infect people) on surfaces for up to four weeks, and you can shed infectious virus for 72 hrs after symptoms end.
- Hand sanitizer and Lysol wipes don’t kill it. Alcohol doesn’t kill it.
- Soap and water followed by fresh bleach in a freshly made solution or steam cleaning (or hot laundry with oxygen or chlorine bleach) works.
With the reduced size of the air inlet, that sped up the air velocity going in and got an anemometer reading of 1.9-2.5 mph, which is great!
Father-in-law is a structural engineer and he volunteered to do a rough CFM calculation based on that and the area of the inlet-- 334 cfm! The max based on the fan specs is 360 cfm, so it seems like a reasonable ballpark. Other testers have found these filters to be about 85% efficient, so the CADR cfm should be in the 240-280 range.
Couple more pictures of adding a cardboard shroud/shade and construction paper sides.
The shroud shifts the air inlet to the bottom, near to the table. Prior to that, the most airflow was at the top of the filter.
It also acts as a shade, directing the light through the filter down onto the table.
psychopomp disaster
Dominik Moritz
fritzoids
Clive Thompson
Erin Kissane
arnicas