ks2048

Walmart Logo

I was put off by the vibed-design, but everything looks well done and well-explained.

From what I can tell, it used 5.3 hours of single voice fine-tune data.

It was based on an existing TTS, IndicT5. I wonder how different is “Sanskrit Chanting” to languages it could already do, like Hindi. Is it largely the glyph—to-phoneme that needs relearning? Or pitch control? Or more?

You need some visual feedback that it's loading. I see a blank screen for 30 seconds.

Summary (at end of PDF):

As discussed at the beginning of this article, the excitement with AI is carrying us along in a big wave, but the practitioners whose job it is to make this all work are scrambling behind the scenes, often more in dread than excitement. In some cases, they are using outdated techniques; in others, approaches that only work for now; and every so often they are doing nothing at all in order to meet significant operational, technical, and business challenges.

In MLOps terms, it sometimes feels that we are using older paradigms to manage a thoroughly new situation, and it’s not entirely clear that we really see it like this. We should be casting about for either a better paradigm or a better patching-up of the existing paradigms than is available today. Regardless, we hope that the summary of the problems presented here is a useful stimulant to people attempting to think about them more holistically and, hopefully, helps to provide some answers.

The article being discussed (with reconstructed glyph drawings and description):

https://www.cambridge.org/core/journals/antiquity/article/id...

(PDF button not working for me, but looks like entire contents are here in HTML).

Amazing work and historical artifacts.

Something about this era - I have an interest in Frederick Catherwood and his work at basically the same time in mesoamerica (although he focused more on ruins than modern people), https://en.wikipedia.org/wiki/Frederick_Catherwood

Probably not worth the effort (or legal trouble), unless you can show it's better than other recent open models like Cohere Transcribe.

Lots of comments are "you should compare against X and Y" - even better, just get the results on a standard benchmark, so you can compare against all,

https://huggingface.co/spaces/hf-audio/open_asr_leaderboard

It took a bit of hacking, but most of the data was from the Eurail/Interrail Planner app. I used that book the tickets and it had an HTML export feature - showing your route on a web page - and I stripped the route from that.

Did a trip like this a few years ago. Highly recommended, if you have the opportunity!

Wanted to do a write-up like this, but only got as far as the map,

https://kenschutte.com/europe-2023/

I wonder if they took the new "Rail Baltica" through the Baltics? That was one of weaker links in the train route - I used a bus between Vilnius and Riga.

Even if all her donations were a complete waste, one can argue every $1B taken from the oligarch class is a win - less money to buy media outlets and politicians.

That's what I gathered from the blog post - which made the title of the blog seem odd.

What I don’t like is two things. One, this constant bullshit about some window closing, or the perpetual underclass, or falling hopelessly behind.

The blog has a tagline, "the singularity is nearer". I think belief in a "singularity" almost implies these things to some degree.

LLMs don't notice this. Gray-on-black-small-letters. An immediate close the tab for me.

This seems to be based on Google's QuickDraw datasets. 50 million samples are available in an open dataset,

https://github.com/googlecreativelab/quickdraw-dataset

I agree one man's "freedom" is another man's annoyance (or worse). It is a difficult, fine-line to draw.

Thus, the beer example was mocking the author's view of "you have freedom or you don't" and the simplistic idea that "America is a _free country_".

I, for one, think the ChatGPT response to, "Hey, I just killed my wife ..." is not bad and preferable to it helping him.

If you design a hammer to kill people and then give it away, you are partially liable for the deaths it causes.

This doesn't seem to be the case today (weapons manufacture).

You can criticize being wrong, but why is the doomer argument "misanthropic" or "malevolent"?

Like we either live in a world with freedom or we don’t, and like many Americans who have come before, I’m willing to give my life to fighting for it.

This is a very simplistic view. "Freedom" isn't binary.

In most of "land of the free", I can't even sit on a park bench and drink a can of beer.

Yes, this is just a small example of a personal freedom - and not an important, cherished freedom like his examples (freedom to have a robot help you cover-up a murder).

It seems to hinge a lot on what is “culture”.

This kind of belief should make one stop and think about one's information diet.

I'm also certain that without ICE there would literally be unlimited immigration

ICE was created in 2002.

Blog post idea:

We made Grok 4.5, GPT-5.5, and Claude write a blog post about using Grok 4.5, GPT-5.5, and Claude to build the same apps.

And ICU uses data from CLDR, which is mentioned in the blog. Here, there are 380 xml files: https://github.com/unicode-org/cldr/tree/main/common/transfo...

Yes, ICU is ubiquitous. But, some NLP projects use various other libraries, such as uroman (just for romanization - to Latin script).

I could be wrong, but I don't think it's common for websites to just transliterate any text they're given. Let's check: ウィキペディア

Does the Latin-Katakana example given imply that some input value can cause it to not terminate?

Also will need a pretty big napkin,

https://raw.githubusercontent.com/adriancable/eternal/refs/h...

And Wikipedia says this one is over 12,000m deep,

https://en.wikipedia.org/wiki/Al_Shaheen_Oil_Field