factorymoo

Can anyone tell me if there are similar risks installing software using Brew on macos? I would imagine so.

Hi Peter, thanks for doing this AMA!

I’m a cofounder of a startup in the US. Two of us are here on green cards, but our third cofounder is based in Switzerland. He has a PhD from a top university, previously founded a company, and has raised over $30M in the past.

At what stage would it be possible for us to bring him to the US on a visa? Would it be:

- As soon as we incorporate a C Corp?

- After raising funding?

- Once we have revenue?

Are there any specific visa pathways (O-1, L-1, E-2, etc.) that would be most relevant for him, given his background?

Appreciate any guidance on this!

To me less boring. I used to struggle to understand new concepts as they were presented. That year though, I was able to follow what the teacher was saying "live", ask interesting questions to deepen my knowledge.

I went to the most prestigious high school in France. The top 2 students in my maths class shared one thing in common: they would study the curriculum the summer before.

I did it one summer, and while I was nowhere near as good as them - something magical happened: even though I hadn't understood all the concepts, my ability to understand the concepts during the class went way up. It was easier to follow what the teacher was saying since no concept was totally new to my mind.

Why does Google even work on this type of problem? I am genuinely at a loss here.

Same edit: actually it takes like 5+ minutes

your property probably goes up in value.

decreases the cost of shelter.

Which is it?

I work in Big Tech. Whenever we open positions both in the Bay and in NYC, we are flooded with applications to NYC with a lot less in the Bay. A good chunk of these come from people in the Bay wanting to move out.

I've been in the Bay for a few years now. I've noticed that a lot of people I talk to don't really like it here. They like their job and the paycheck but they would move out in a heartbeat if they could. As opposed to all the folks I know who live in NYC, most of them really enjoy it.

I wonder if that has something to do with it ["it" being the article]

I asked it for an efficient way to sort a list in Python [1].

I'm running the code it gave me to try it out on a small list, it's been 10 minutes and it's still running. Might be something worth looking into.

Granted, the way I asked for this function was not the most natural.

[1] https://devv.ai/search?threadId=dl4c8if11c00

That's fair, though I would argue that you have more impact within the organisation, but not necessarily more impact on more users.

This might be an unfair statement but it really feels like all of these blogs don't know why. They copy/paste each other (you often seem the same errors in multiple notebooks/blogs) and I have a feeling no one really deeply understands what they're doing.

Yeah good point - that would probably be ok for me actually.

I would argue that a core benefit of using Substack, beyond the easiness to write, publish, send emails, is trust for payments. I would be more reluctant at putting my credit card info in a random self hosted blog vs an established company's like Substack.

Potential solution: download all videos locally [1] As an added benefit, you'll skip all sponsored content.

Just put this in your terminal once you install yt-dlp:

yt-dlp --sponsorblock-remove all URL "put a youtube URL here without quotes"

[1] https://github.com/yt-dlp/yt-dlp

As someone who has worked at big social media companies, I can say the conventional wisdom about content moderation being a carefully planned process is off-base.

The reality is that moderation relies heavily on imperfect machine learning models and overworked human reviewers making rushed judgments on hundreds of cases per day. There's no meticulous strategy document mapping out the pros and cons before banning accounts that upset the company.

Mistakes inevitably happen when relying on this combination of flawed automation and human reviewers who are stretched too thin. The moderation policies may seem arbitrary or politically motivated from the outside, but much of it comes down to hasty human error and buggy algorithms rather than some malicious scheme.

I can't seem to find a link to the git or documentation. I'm very curious how this is achieved? Is a model fine tuned? If so how? I'd be most interested in seeing how they formatted the input data if fine tune was done.

" with zero ability to think through long-term consequences"

The median tenure in tech is one to two years before moving on to another team or company [1]. You need to say that what you care about is long term, but that's now how we're compensated (read: incentivised). Plus you're not there to see it anyways so there are really very little incentives to think long term.

[1] some googling but couldn't find a great source for this. Though it matches what I've observed in the industry.

As a new (ish) manager, curious to hear more about your style and what has worked well for you for 25 years.

Interesting to see the perspectives from early hacking pioneers. Seems like some things haven't changed much - debates over ethics of unauthorized access, whether it's criminal, free speech implications, etc. But more nuance now as hacking's gone more mainstream.

Biggest change is probably threat models. In 1990 main concern was individuals hacking systems for challenge, curiosity, etc. Today it's nation-states and organized crime using hacking for financial gain, espionage, even kinetic attacks.

Other change is commercialization/professionalization of hacking. Now huge industry around cybersecurity, ethical hacking, bug bounties. Hacking skills lead to lucrative careers, not just hobby or activism.

More diversity today too - no longer just male techies. But part of cyberpunk spirit remains, even as hacking's become bigger business and political issue.

Reminds of me a site I used when house hunting:https://www.mapnificent.net/

I worked for Amazon for two years. They always were very proud of their customer obsession and in some ways it really showed (great customer service, very speedy deliveries).

But what I never understood is how it never seemed to be a problem that over the years, the site is just filled with low quality / alibaba imports / bad products.

On a side note, the number of vendors that offer you to remove a bad review in exchange for a full refund (and you get to keep the product) is also making me lose all trust in their review system.

Product idea: use reddit to obtain 3 price point options for all items deemed of high quality (or at least that people are happy with). Say cordless vacuum cleaners and have an extension in chrome that only displays these 3 options when you look up for this item in Amazon.

Same question for Uber with 36k employees. It's an app that connects taxis to clients. I guess I'm showing a lot of ignorance here - but why does this need 36k people?

I guess this was Elon's thoughts when he laid off most of twitter.

Edit: Found a good answer for Uber on quora: https://qr.ae/py4zBU

"Transformers" and "Attention is All You Need" refer to an important development in machine learning and artificial intelligence, particularly in the field of natural language processing (NLP). I'll try to explain them in a simple way.

Think of a conversation you had with a friend. While they were talking, you were probably not just listening to the words they were saying right now, but also remembering what they said a few minutes ago. Your brain was connecting the dots between different parts of the conversation to understand the full meaning. Now, imagine if you could only understand each word in isolation and couldn't remember anything from a few seconds ago. Conversations would be pretty hard to understand, right?

In early NLP models, this was a big problem. They couldn't easily look at the "context" of a conversation or a sentence. They could only look at a few words at a time, so they were a bit like our forgetful person. They were good at understanding the meaning of individual words, but not so good at understanding how those words fit together to create meaning.

Linux n00b over here... apologies for the silly question but I can't seem to make this work

I enter spongebob-cli

Then when it prompts me to enter an episode number, I can't seem to get it to play from within the terminal itself

let's say I want to play episode 1:

1 --mpv --vo=tct --quiet

Doesn't seem to work. Any help to get past this would be greatly appreciated

As of Firefox 93.0

From the release notes: "Firefox PDF viewer now supports filling more forms (XFA-based forms, used by multiple governments and banks)."

https://support.mozilla.org/en-US/kb/view-pdf-files-firefox-...