subtypefiddler

"Amazon representatives were invited to attend an employment committee hearing on "Working Conditions in Amazon Warehouses" on Jan. 23, but did not show up, claiming that the notice of about a month had been too short."

“It is unreasonable for members to be lobbied by Amazon while at the same time being deprived of the right to represent the interests of European citizens and inquire about claims of breaches of fundamental rights enshrined in EU Treaties and EU labor laws”

The ban seems very reasonable!

It boils down to

- Labor relations (unions in Arizona pushed back agains Taiwanese workers build the factory)

- Local partners (Denso/Sony and Toyota investing in Japanese project, TSMC on its own in the US)

- Subsidies (Japan delivered on promises, US didn't)

- Ambition (12nm-28nm in Japan, 4nm in US)

It seems the US gov is not very serious about it while Japanese gov surely is. It sounds self-inflicted.

(edit: formatting)

Anecdotal evidence from France and UK doesn't agree with "isn't generally true outside of the US". When a diverse group of people is involved it's way more likely to use WhatsApp (or Signal to a lesser extent) outside the US, but iMessage is still used a lot IME.

I don't think it's a new form of discrimination either, brand awareness and social position via brands is incredibly prevalent in teenagers and has been for a while.

It's not a great piece from FT.

It argues that "the root of maths chauvinism is a childlike craving for certainties, or at least probabilities, amid the flux of adult experience", which I kind of agree with. Certainty is indeed one of the appeal of mathematics.

Using Liz Truss as an example - she studied PPE at Oxford, not exactly a math heavy major - is beyond weird.

I don't understand the broader point Ganesh is trying to make.

Very interesting examples. For Nash, Von Neumann contributions to the field were orders of magnitude more important. For Kahneman and Tversky let's not forget that the foundations are quite shaky to quote [0]

[0] https://replicationindex.com/2020/12/30/a-meta-scientific-pe...

She wrote a paper on blind auditions, this gained a lot of media attention, but as a lot of her work it was quite exaggerated see e.g [0]

[0] https://statmodeling.stat.columbia.edu/2019/05/11/did-blind-...

At first I was surprised the economist would talk about the networking stack, but this is about social networking. The introverts bit makes much more sense.

Regarding the article it says "because they bring you new information, more infrequent and distant relationships (or “weak ties”) are more useful than close contacts." Mark Granovetter studied that in 1973 in The Strength of Weak Ties.

There's been a lot of community backlash in Pokemon Go (#HearUsNiantic Niantic movement)and the game director response to it was that the company has “no plans to directly address any of [it]” [0]

The fan base is very loyal to the Pokemon brand and it really feels Pokemon Go is successful despite all of Niantic's shenanigans.

[0] https://www.reddit.com/r/pokemongo/comments/13lfd1c/its_offi...

It's all WebKit on iOS and iPadOS

It's just The Times, not "The Times of London"

The Times != The New York Times

Looks really good! Maybe add in the title that it's MacOS only for now. There's options for local and cloud-based AI - which is nice. I'd be cool to see what's the compute cost of requests!

Is it true though? Doesn't seem to be the case for healthcare outcomes [0], vacation time [1], labor rights, parental leave, etc. [0] https://pubmed.ncbi.nlm.nih.gov/33369633/ [1] https://en.wikipedia.org/wiki/List_of_minimum_annual_leave_b...

Well, it's a great example as Dr. Earl Haas patented the first tampons in the early 30s and nobody cared. He then sold the patent to a business woman that made it successful. So effectively a man invented it but failed to market it and sell it well.

"Funding/Support: This study was partly funded by the CCSA as a subcontract for a Health Canada grant to develop guidance for Canadians on alcohol and health."

BUT

"Conflict of Interest Disclosures: Dr Stockwell reported receiving personal fees from Ontario Public Servants Employees Union for expert witness testimony and personal fees from Alko outside the submitted work. Dr Sherk reported receiving grants from Canadian Centre on Substance Use and Addiction (CCSA) during the conduct of the study. No other disclosures were reported."

Where "Alko Inc is the national alcoholic beverage retailing monopoly in Finland."

> "Of course adding poison to your body is unhealthy and will affect mortality."

This is in contradiction with the basic principle of toxicology: "dosis sola facit venenum".

A "systematic review and meta-analysis of 107 cohort studies involving more than 4.8 million participants" cannot simply be shrugged away the way you do.

In the case of municipal broadband it mostly comes from communities that suffer form poor/no investment from the big players and is the only recourse for decent internet. It usually is faster and cheaper than existing solutions. There's clearly market failure here and players abusing their local monopolies to refrain from investing / provide decent service. The state of telco in the US is pretty shocking, even in places like NY state.

> study that isn't peer reviewed

Terrier, C., 2020. Boys lag behind: How teachers’ gender biases affect student achievement. Economics of Education Review, 77, p.101981.

"Economics of Education Review is a quarterly peer-reviewed academic journal covering education economics. It was established in 1981 and is published by Elsevier."

What are you talking about?

It's also important to note that they work despite being wide, you can see that with the efficiency of pruning, and ideas such as the lottery ticket hypothesis that state that "successful" sub-networks within the wide network account for most of the performance.

In the theory literature, if you have a K-deep network, K=1 is the shallow case, K>1 is deep. Agreed naming could be better, but it's not like "deep work" or "deep thoughts" as the parent was stating.

Whilst I agree with the general sentiment, in this particular instance it has to do with the depth of network that could be trained efficiently thanks to hardware advances. LeNet was 7 layers deep, Dan's 9, VGG's 13, GoogleNet's 22, etc.

There is theory w.r.t to thick networks as well (e.g the link to Gaussian processes require infinite width).

Deep makes sense here.

The dichotomy between Greenwald's complaints (censorship of his article despite contractual guarantees, Reality Winner cover-up, what editors forced Lee Fang to do, lack of reporting of Assange hearing, and "lack of editorial standards when it comes to viewpoints or reporting that flatter the beliefs of its liberal base") and the editor in chief response in the NYT [0] (he is a "grown person throwing a tantrum") is frightening.

[0] https://www.nytimes.com/2020/10/29/business/glenn-greenwald-...

Full disclosure, European here and in our team everyone is found of Linux so not representative of your average MLEngineer.

We actually had more issues with nvidia drivers messing up newcomers' machines during updates than with setting up AMD GPUs, but then again n is small (and AMD GPUs were for playing around rather than real work).

Still, a Titan Xp has CUDA support and plenty of memory, but it's better, IME, to upgrade to a model with less memory but higher cuda compute and access to tensor cores.

It's not just CUDA vs ROCm, ROCm has come a long way and is pretty compelling right now. What they lack is the proper hardware acceleration (e.g tensor cores).

Links to avoid verizon's obnoxious "policy-wall":

- https://outline.com/XZYvKa

- https://web.archive.org/web/20201026181132/https://techcrunc... (seems to successfully load then fails after a few seconds)

When reading the source https://my.visme.co/view/01070kx3-unconventional-ventures-no... the divide goes:

Gender | % Companies (#) | % Funding

All men | 85 (1473) | 88.81

All women| 6.35 (110) | 2.21

Mixed | 8.66 (150) | 8.98

Notes:

- There is no disparity when it comes to number of rounds

- The trend is decrease in funding from all men teams (-10 points compared to 2019) and increase mostly for mixed teams (+10 points) and women (+2 points) (their numbers don't add up to 100...)

The article reads "On top of the female-founded startups finding it difficult to raise any money at all, the 6% of startups that did, only received 1.3% of the funding.".

I understand it as 6% of female-founded startups managed to raise money and they received 1.3% of total founding, rather than 6% of startups were all-female.

The article doesn't give the proportion of mixed and all-female teams, just of all male teams. To quote you "make an effort".