cschmidt

The fashion houses all ended up looking exactly the same. Burberry is the first to swing back the other way, returning to an updated serif logo.

https://twitter.com/ucllc/status/1622979476899393536

So, hopefully the identical logo trend has peaked.

One level 50 problem was Exercise 1.2.6.63:

"Develop computer programs for simplifying sums that involve binomial coefficients"

Which was solved in the book titled "A = B":

https://www2.math.upenn.edu/~wilf/AeqB.html

or

https://www.amazon.com/B-Marko-Petkovsek/dp/1568810636

I just asked GPT-4, and it denies being a MoE model.

User:

Are you an MoE model?

ChatGPT:

No, I am not based on a Mixture of Experts (MoE) model. My underlying architecture is based on the GPT (Generative Pre-trained Transformer) framework, specifically the GPT-4 version. This architecture is a large-scale transformer-based neural network, but it does not use the MoE approach.

In a GPT model like mine, the entire model works in a unified manner on each task, without dividing it among specialized sub-models or experts. The model is trained on a diverse range of data and tasks, enabling it to generate responses across a wide array of topics and questions. While this approach lacks the specialization that MoE models offer, it provides broad, general-purpose language understanding and generation capabilities.

The Economist also wrote an article about this last week, spurred by the class action lawsuit. Paywalled as well, but I'm hoping you have some free articles left, unlike the WSJ:

https://www.economist.com/leaders/2023/11/08/time-to-take-a-...

If you want to know more about Paul Rand, who is generally considered the greatest logo designer ever, I recommend:

https://www.amazon.com/Paul-Rand-Steven-Heller/dp/0714839949...

And his own books on design are all worth a read.

He did the classic logos of IBM, NeXT, UPS, Morningstar, Cummins Engine, CBS, Westinghouse, and lots more.

This probably doesn't help you directly, but I suspect the older people mentioned in the article may often have some sleep apnea, preventing them from reaching deep sleep. I just started using a CPAP machine, which has greatly improved it for me. Making sure it is always treated could help society wide.

As I mentioned in another comment, they voted down ranked choice voting in Massachusetts where I live, which really made me sad.

In the next election there was an open seat for my Congress person. There were something like 5 progressive candidates, and one more conservative democrat. Of course, the progressive vote split and the conservative candidate won, even though he was probably the last choice among a majority of voters. I found that super ironic.

I don't think American voters could understand that very well. Where I live in Massachusetts they defeated a ballot initiative for ranked choice voting, which is just clearly better (in that it reflects the will of the people), especially in a state dominated by a single party.

That work I was referring to is a creative, artistic process of designing a font based on a historical starting point. That does matter. Hardly mowing the lawn or compiling a phone book.

Historical revivals are hardly the same thing. They often end up far from the original, and require a ton of work. It isn't just a matter of digitization. And they "claim copyright" on their derived work, not the original. With fonts they really just copyright the name.

I also read it recently, and didn't care for it much. But I was always more of a Asimov / Clarke fan than of Heinlein, so I wasn't super surprised.

I see LaTeX as the only game in town if you're writing a paper in math / CS / engineering / etc. The conference or journal provides a formatting template, and you use that. These days it seems like Overleaf is very commonly used to provide collaboration between authors. So it isn't really a choice, despite all the flaws.

> Roon is the only other music player that understands the difference between compositions and performances like Apple Music Classical does.

Idagio mentioned in this thread does this as well. You can choose your performance by conductor / soloist / ensemble etc.

Yes I think you’re right. I was talking about the standalone enter key, or you could also do some modifier and return. It has been a long time

Yes. If you ever find yourself buying a webfont for your latest website or logo, always see if you can buy it direct from the designer. Monotype takes a huge (like >50%) cut if I remember correctly.

For example, my last project I used MD System from Mass Driver for the web fonts (https://mass-driver.com/typefaces/md-system) and Denton from Peregrin Studio (https://peregrinstudio.com/work/denton) for my logo.

I think it was just the enter key to execute the selected text.

This paper is focused on the little prizes. The odds of winning the jackpot will be the same for any play.

In fact the only thing you can do in the lottery is to avoid "popular" sequences like birthdays etc, so that if you do win you don't have to split it with a bunch of people.

A lottery in the Philippines famously had 433 winners, who all played 9, 18, 27, 36, 45 and 54. (https://www.cnn.com/2022/10/06/asia/philippines-grand-lotto-...)

While on the subject of ACL 2023 tutorials, here are two I went to:

Complex Reasoning in Natural Language https://wenting-zhao.github.io/complex-reasoning-tutorial/

Generating Text from Language Models https://rycolab.io/classes/acl-2023-tutorial/

the second one in particular was very well taught. I didn't catch the Retrieval based talk, since it conflicted with the text generation one.

The Pile and Red Pajama are primarily English language datasets. If you want something multilingual, I'd suggest having a look at the Bloom dataset https://arxiv.org/abs/2210.14712

If you’re looking at The Pile, you also might consider the Red Pajama dataset. A new cleaned version was released recently https://www.cerebras.net/blog/slimpajama-a-627b-token-cleane...

The python version is great news. I get asked fairly frequently to recommend an intro ML book. I would have suggested this, except they usually only knew python and not R. Now it a perfect first book!

Cory's article about enshittification of TikTok, for those that don't like audio

https://pluralistic.net/2023/01/21/potemkin-ai/#hey-guys

Here's an interesting Economist article on the same subject:

https://www.economist.com/science-and-technology/2023/06/07/...

Last week the Economist had several stories from an 8 hour interview they did with Kissinger. Interesting stuff...

https://www.economist.com/briefing/2023/05/17/henry-kissinge... https://www.economist.com/podcasts/2023/05/20/henry-kissinge... https://www.economist.com/kissinger-transcript

They are emulating different HP calculators. The DM42 is a HP-42S clone, while the OP is a HP-32SII.

Another interesting article on the topic here:

https://tedgioia.substack.com/p/a-scandal-is-taking-place-in...

Finally replaced this week!

https://www.underconsideration.com/brandnew/archives/new_log...

The Economist has a whole series of this type of article, basically what to read to learn about X.

https://www.economist.com/the-economist-reads

You're right, it has changed since this morning.

I'd like that as well. For a while I thought https://jazzed.com/ would eventually be this, even thought it was just available in the UK. But now they seem like they've moved to focusing on live performances.