[thread]

I am going to do a thread on several numbers

1

2,996

4,494

43,730

106,158

1/ Throughout my startup career, I’ve seen many founders have an aversion to sales. This is a mistake that can kill your startup.

2/ Sales is often viewed as either a saving grace or proof that the product isn’t good enough (because it should sell itself). Neither are ever true. Some common mistakes that result in...

3/ Mistake 1: Hire a sales rep before reaching product/market fit to get your initial batch of customers. This is a mistake because founders need to work through their MVP with early adopters to truly understand what it is they’re selling.

4/ Mistake 2: Reach product/market fit, need to scale, and rely entirely on self-serve. For enterprise products that require big commitments and internal shifts, almost no product is self-explanatory enough to sell itself.

5/ Mistake 3: Make a first sales hire who isn’t scrappy enough to help mold the sales process from scratch. Some salespeople are amazing at their jobs, but not cut out to establish the processes that others end up following. This skillset is what @rdedatta calls a “sales ninja”.

1 of 16: WE ARE ALL JEWS HERE: THE STORY OF RODDIE EDMONDS

One of the most moving and relevant stories of the Battle of the Bulge, or any American Soldier in any war, is that of Master Sergeant Roddie Edmonds, a Knoxville, Tennessee native, who served with the 106th Infantry.

2 of 16:

Roddie was captured early on in the Battle of the Bulge, on December 19th, when Panzer forces plowed through his unit.

He, along with almost his entire regiment, was forced to surrender.

3 of 16:

The men were transported to the Stalag IX-A POW camp in Ziegenhain, Germany.

Roddie was the senior enlisted American Soldier at the site. As such, he was the conduit between all American Soldiers and their German captors.

4 of 16:

In late January, the camp’s commandant, Major Siegmann [pictured here], ordered Roddie, a Christian, to identify all Jewish Soldiers and order them to stand in formation by themselves the next day.

[Jewish Soldiers were a minority within American units]

5 of 16:

Throughout WWII, captured Jewish Soldiers were often tortured and then killed by their German captors.

In fact, Jewish Soldiers had been told to bury their dog tags before capture. The dog tag identified Jews with the letter “H” for Hebrew.

One fascist to watch in 2021 is Joseph Charles "Oakman" Golinske of SE MI. He's been a Proud Boy for a couple years, but took a break recently from the group and promotes himself through the "Oak Nation" channel on Telegram.

Most people may be familiar with him as the tall guy in the Proud Boys bridge shot, from their August 17, 2019 attack on Portland, which featured an American Guard bus trying (and failing) to bludgeon antifascists with hammers.

He was also with his g/f at the time (Dominique Tota, left of Tarrio), and that's definitely a twitter account folks will want to block/report because he posts on it.

He's been plenty active at the pro-Covid/anti-mask/ReOpen protests in Lansing MI, seen here with Jerry Wayne and an effigy of Gretchen Whitmer, who his militia friends tried to kidnap and publicly execute.

Outside of the Proud Boys, his solo "career" with Oak Nation features the same misogyny, classism and gateway to far-right domestic terrorism the Proud Boys themselves feature. Here's him peddling oxychlorinque and bragging about harassing service workers.

Important paper from Google on large batch optimization. They do impressively careful experiments measuring # iterations needed to achieve target validation error at various batch sizes. The main "surprise" is the lack of surprises. [thread]

https://t.co/7QIx5CFdfJ

The paper is a good example of lots of elements of good experimental design. They validate their metric by showing lots of variants give consistent results. They tune hyperparamters separately for each condition, check that optimum isn't at the endpoints, and measure sensitivity.

They have separate experiments where the hold fixed # iterations and # epochs, which (as they explain) measure very different things. They avoid confounds, such as batch norm's artificial dependence between batch size and regularization strength.

When the experiments are done carefully enough, the results are remarkably consistent between different datasets and architectures. Qualitatively, MNIST behaves just like ImageNet.

Importantly, they don't find any evidence for a "sharp/flat optima" effect whereby better optimization leads to worse final results. They have a good discussion of experimental artifacts/confounds in past papers where such effects were reported.

1/
Hello, my name is Gal,
I am an Israeli citizen, and I want to convey this urgent message to the world: Contrary to what you are told, Israel has not "returned to normal" in any way >>

2/ Moreover, despite the desire to create a setting as if there was a full and normal routine here - the Israeli Ministry of Health JUST RECOMMENDED TODAY TO EXTENT THE DECLARATION OF A “STATE OF EMERGENCY” in Israel >>

3/ to which is the foundation for coercion, violation of basic rights, severe discrimination under a "green mark (pass)", prolonged closure (Of about 5 months overall) and a lot of undemocratic and not “routine" restrictions. >

4/ If the situation in Israel had "returned to routine thanks to the astounding success of vaccines" then the "state of emergency" would not have continued.

5/ In fact, since the beginning of the "vaccination project" there has been only more intimidation and in many ways - even though the closure has been lifted, much more coercion.