polynomial

NFC inlays are paper thin (~0.06mm) and cards are typically 0.2mm (Bicyle) to 0.30mm (M:TG).

We can use NFC tools to write an URI pointing at an audio file link using NDEF.

I believe Android will play the audio file when you tap the card on the your phone. (Apple will need you to confirm in a popup.)

So much for Difference and Repetition.

Evals don't fix this.

Insisting on having a SPF (single point of failure) for... reasons.

So basically, leadership is coping? No thanks.

Sure there is.

The article says the victim used 2fa. How did the attacker know their 2fa in order to send them a fake 2fa request?

Serious question, how did the attacking site (npmjs.help) know the victim's 2fa? ie. How did they know what phone number to send the 2fa request to?

Do you have any citations for various estimates? This is super interesting to me.

Photoshop was hella fun, turned out that programming paid more. And now AI pays much more.

"Realism is overrated." /s

I still feel like most humans would fail, haha.

Or was that just marketing?

Asking the important questions.

Why was it named that is also a question.

New startup idea just dropped.

We don't go after moving violations anymore (in NYC) because the driver might have a bad reaction. True story.

Imagine if you had to do this with an actual team member.

Apparently they have made HUGE breakthroughs in Online Bin Packing.

Extremely Bullish.

He might be a murderer, many such cases.

What phase of the AOL lifecycle is Meta currently in? (Seriously asking.)

At first I read this as "My 2.5 year old can write Space Invaders in JavaScript now"

Persevered through the article and comments in hopes someone would point this out.

"There seems to be some confusion" could literally be Google AI's official slogan.

> learning new key bindings

Why are they shipping them with different key bindings? Seems like the opposite of what you do to encourage product adoption.

Just to note that Tiktoken is still the tokenizer behind the GPT-4x series, it just uses a different token model. (Post only says GPT-3, implying they were using something else for subsequent iterations.)

It's a popular opinion these days, all the cool kids are doing it

I thought best guesses were that Claude's system prompt ran to tens of thousands of tokens, with figures like 30,000 tokens being bandied about.

But the documentation page linked here doesn't bear that out. In fact the Claude 3.7 system prompt on this page clocks in at significantly less than 4,000 tokens.

Is that really supposed to be Zuck with an older version of himself??

let me know when you're doing a standup set and I'll be there.