ramzyo

Is this what you mean? https://chatgpt.com/share/6895632c-fb58-800e-b287-b7a98ad64d...

Hmm, lived there for a long time and walked by Boston City Hall almost every day. I'm not quite sure how to differentiate objective beauty and subjective beauty, but at least subjectively, in my opinion, it's an eyesore.

Agreed on all fronts that City Hall Plaza is a disaster, though. I thought there were plans to revamp it with the Government Center station green line revamp a few years ago, but not sure if that improved anything.

Exactly this. Have heard it referred to as "break glass access". Some form of remote access, be it SSH or otherwise, in case of serious emergency.

Frigate uses the TPU with an application processor, not this MCU + TPU version, right?

Yeah, I didn't understand the purpose of this device when it came out and still don't. It's an interesting system but seems like worst of both worlds because the coral TPU and M7 aren’t low power enough for battery applications, and it's unclear whether the full 4 TOPS of the Coral is achievable given the MCU’s memory bus bandwidth. So to me it looks like a computationally underpowered system that you have to keep plugged into the wall.

Going with the Cortex A Coral Dev Board or another SBC with the PCIe or USB standalone Coral TPU seems like a better bet. You'd get a better camera (eg via USB), more processing power and memory, and more full featured software (both Linux and TFLite instead of baremetal or embedded OS and TFLite Micro). Price point would be higher for this option, but you'd certainly make that up in saved time very quickly not having to deal with baremetal programming or an embedded OS.

Vendoring means to bundle dependencies into a project directly (often this means copying source code and using/maintaining the copy) rather than from some other source (other packages, OS, package repo, etc). Here's an article from LWN that talks about it with a real world example: https://lwn.net/Articles/842319/

I think this maybe could indicate potential increased usage of reverse proxies? Total shot in the dark though, not sure.

That's right, origin and destination are still visible. Even if you use encrypted DNS to hide hostname to IP lookups, your actual traffic has to be routed somehow by someone. Whether that's your ISP or a VPN provider + their ISP.

Posted this on another thread about VPNs a few weeks ago. Reposting here since I think it applies.

I've recently been describing what a commercial VPN provides to non-technical friends and family as a type of "global virtual Internet cafe" subscription - the pros and cons of using a physical Internet cafe mostly apply. An Internet cafe isn't inherently (i.e. due to technical benefits of underlying technology) any more or less secure than connecting to your home or work wifi/network, and the Internet cafe knows who you are and what websites you're visiting, but your ISP/employer doesn't (since you're "at" the Internet cafe, not on your home/work network).

Of course, your ISP/employer does know that you're visiting the Internet cafe, and in the case of work (and some ISPs) can stop you from doing so.

If you visit a website from an Internet cafe, the website may still be able to figure out who you are, just like they can when you bounce between different networks normally. And of course, if you login to your account on a website or put your shipping address or something in when buying something, you're self identifying (unless you have throwaway accounts or forwarding addresses or whatever).

And finally, if someone really wants to figure out who you are to a high degree of confidence, they will.

I find this lands pretty well and is close enough to being technically correct without getting into the details that non-technical people would start glazing over if I got into.

I've recently been describing what a commercial VPN provides to non-technical friends and family as a type of "global virtual Internet cafe" subscription. It's not inherently (i.e. due to technical benefits of underlying technology) any more or less secure than connecting to your home or work wifi/network, and the Internet cafe knows who you are and what websites you're visiting, but your ISP/employer doesn't (since you're "at" the Internet cafe, not on your home/work network).

Of course, your ISP/employer does know that you're visiting the Internet cafe, and in the case of work (and some ISPs) can stop you from doing so.

If you visit a website from an Internet cafe, the website may still be able to figure out who you are, just like they can when you bounce between different networks normally. And of course, if you login to your account on a website or put your shipping address or something in when buying something, you're self identifying (unless you have throwaway accounts or forwarding addresses or whatever). And finally, if someone really wants to figure out who you are to a high degree of confidence, they will.

I find this lands pretty well and is close enough to being technically correct without getting into the details that non-technical people would start glazing over if I got into.

I agree for casual browsing, but just curious for any service that requires providing some other piece of identifiable info (eg a site you have an account with and login to, or buying something online and entering your delivery address), what is a VPN helping with other than hiding that activity from your ISP (which may be what you're going for, but just curious)? Once you identify yourself to a website or online service via typical means, how is the IP anonymity a VPN provides helping?

Looks like they had a meaningful conversation about it a little later. https://youtu.be/jstD4tIuRuI

Yeah, I agree that’s pretty confusing. That being said I had to read the Wordle rules the first time I played. Had no idea what the differences were between colors and that a letter could be used multiple times in the same word.

Note that Neural Magic's engine is completely closed-source, and their python library communicates with the Neural Magic servers. If you use their engine, your project survives only as long as Neural Magic's business survives.

Could be totally off here, but have a feeling this team is going to get scooped up by one of the big co's. Having deja vu of XNOR.ai (purchased by Apple), when partners got the short end of the stick (see Wyze cam saga: https://www.theverge.com/2019/11/27/20985527/wyze-person-det...)

This is really impressive work. Would be interested to know whether they plan to support ARM CPUs in their runtime engine (currently looks like just a few AMD & Intel CPUs: https://github.com/neuralmagic/deepsparse). Many IoT and embedded devices with application-class processors could benefit from these speedups.

This is a loaded question. "society should value" and "pay more" are separate arguments. By agreeing to the first I'm not agreeing to the second. I think society should value taking time off to raise children, yes. However, I don't think that society should value a corporate attorney doing this more than a public defender doing it. To do so would value raising children unequally. A universal basic income with an additional stipend for child care would be one solution that captures this difference.

Yes, I agree. I don’t see how that ties to the original argument in the thread, though. Presumably a full time parent with a spouse who can provide for both parent and child and who gave up a taxable wage to be a caregiver already had room and board. The room and board comment seems irrelevant to the argument given the context of the preceding comments around forfeiting a taxable wage for the full time job of parenthood.

No, you’re right that doesn’t make sense, and I wouldn’t be for that. I think government should expand paid time off for family leave, especially in the US.

That’s a pretty narrow interpretation of “marriage” and doesn’t track my experience at all.

Effectively they do get paid. In a marriage, mothers own half their husband's income.

That’s some creative logic. Being a stay at home parent is a full-time job. If the stay at home parent were being paid, they’d...well...be paid. “Effective” payment isn’t helpful to a full time parent raising a child and losing out on wages they would otherwise get at a job that the economy values with a taxable wage.

However the author appears to be arguing from a position that they (or someone) actually know how the brain works

Interesting that you had this takeaway, I didn’t get a sense of this at all. My takeaway was that the author presented modern findings (which constitute our present model for how the brain works, speaking nothing to a notion of some absolute correctness) to dispel older hypotheses that are disprovable based on the latest research. As presented, IMO, the author captures well that these modern findings are just a reference point against which to refute the stubborn tropes.

Ah, that clears it up for me. Thanks!

Yes, the author captures this in their analysis as a risk variable.

Interesting take, could definitely happen.

Regarding the rate - LTCG depends on ordinary income level. The 2020 rate for someone earning $150k was 15%, so the $10 million exit in the author’s example would be taxed accordingly. Unless I’m missing something here?

Don’t mention it! Nice article and again kudos to you on the career success.

Well-written article and kudos to the author on their career success!

One nitpick, the part about comparing possible startup comp in a unicorn exit and as a senior eng at Facebook over an 8 year period is missing the possible tax benefits of the startup comp. Facebook comp will be taxed as ordinary income as the stock is granted as RSUs (basically 40 - 50% in the US in the $500k/year bracket depending on the state you live in), whereas the majority of a $10 million windfall from being an early employee at a startup that exits after 8 years would likely result in at least half of the tax burden if you did it right and the options were ISOs. With a 4 year standard vesting schedule, the majority of options would have vested far before the startup was a unicorn, and the tax burden at time of option exercise may be very low. Assuming then a long term capital gains rate of 15% at time of sale, the comparison to Facebook comp over the same timeline is about 4x in favor of the startup rather than the 2x the article proposed.

As the author says, though, the startup route is exceedingly risky and very highly likely not to come out ahead.

Does this exist? If so, what is it? Genuinely curious.

We've run into the same problem. What's made this a particularly hard problem for us to address is that the intentions of involved parties are usually aligned with the best interests of the business in the short term despite being misaligned (and detrimental) over the medium and long term. In discussion it's clear that both parties are doing what they believe is right for the company, but their level and responsibilities might inadvertently encourage them to optimize for the short term. The ensuing conversation is usually a tight rope walk between trying to get both parties to understand short term vs long term outcomes and consequences while encouraging them to continue to execute on what they believe is right for the company.

What's worked for us is to not restrict communication, but to hold our engineers accountable to what they say they'll get done and when. If something slips and it's not for a technical or personal reason (i.e. some risk materializing or needing to take unplanned time off), but rather because someone floated in and asked for something that wasn't scoped in, that comes up in a review and reflects poorly on the engineer and the asking party. On the other hand, if the engineer is able to get all of their scoped work done and also get the unplanned work done, that reflects well on the engineer and potentially the asking party. In our experience, approaching these situations in this way also has the benefit of preserving and encouraging individual autonomy.

This is an informative counter-example to those presented in the article. Thank you for sharing.

It’s interesting you say that, because that’s what I tell people on my team, and I mean it.