Heavy items such as milk goes in first, so that they will also be at the bottom of my backpack. Light and fragile things, such as salad goes last.

So, if I use espresso cups or even dollhouse cups, I can easily hit that optimal cup number with a very small amount of coffee. 😁

Here’s the interesting bit.

“The association between coffee consumption and all-cause mortality was U-shaped; coffee intake seemed optimal at 3–5 cups/d with the lowest risk at 4 cups/d (HR: 0.68, 95% CI: 0.53, 0.88). Our results suggest that coffee consumption may be associated with a lower risk of CRC recurrence and all-cause mortality.”

Well, what exactly counts as a cup then?

“Coffee consumption in cups/d was then calculated by multiplying the frequency of consumption per day by the number of cups that were consumed. We further accounted for the differences in the sizes of cups by multiplying coffee in cups/d by 1 (for cup) or 1.5 (for mug).“

I guess the traditional 100 ml cups and the common 200 ml cups are all just cups, whereas 300 ml mug is clearly a mug. Who knows really, when the data is all self reported. I guess my 400 ml/d consumption should be roughly optimal.

“The truth may be out there, but lies are inside your head.” – Terry Pratchett

Well, that’s exactly what the voting buttons are for.

That would probably freak him out. Imagine that your legacy got turned into a mutated abomination like that.

Reminds me of Youtube. Guns and drugs, don’t make advertiser friendly videos, and that’s why the platform only barely tolerates those videos. From the predictive of TY, they demand resources without giving any direct revenue. I can see the reason why those channels tend struggle so much.

What if the names were picked from whatever news was prominent during the release week. During last summer we could have had something like blackout or APIcalypse.

People who stay there regardless, totally had it coming.

Let that sinking ship join Myspace and Tumblr already. Revolting may rise awareness of the problem, but it isn’t going to change the direction where Reddit is headed. They need to become profitable, and they’ve decided to do that by backstabbing the users.

Closer to 15 years ago. Skype and WhatsApp (before the FB nonsense) were viable options to SMS as long as your friends were also using the same app.

Although, the viability also depended on the price you had to pay for the data. If it’s like 1.5 €/MB, sending snail mail suddenly seems like a very appealing alternative. Some time around 2003-2005 there was still one company that actually charged that much while all the competitors were switching to monthly packages or even unlimited plans. The price range was absolutely wild back then.

A time when AI wasn’t involved with everything.

Remember that time when humans had to do everything, and if there weren’t enough people around to do it, then nobody do it.

Just giving it random data should be good enough. If the users can’t trust the validity of the data, they might stop using the app.

When I was still using Reddit, I used to report those ad posts for terrorism, inappropriate content or whatever term like that.

Many problems have early warning signs. Just train the model to notice those signs.

At least in many industrial applications, the vibration of an electric motor or an axle is a good measurement. Also, the temperature of a ball bearing can tell you a lot. That’s just the basics though, because you can also train the model to look at fancier details.

Three are also tests where you are expected to think like the person who made the test to figure or what the “correct” answer us. It’s not really correct, but it is the one that gets you the points.

Also some IQ question have several correct answers, but only one of them gives you the points. Super annoying. If you’re creative and smart enough to come up with a logically consistent answer you’re still not guaranteed to get the “correct” answer.