rumdonut

They are though. It's that people tend to forget over time. I barely remember details of the last administration. A NYTimes article recently observed the same.

I agree as someone who was previously excited about the idea. Although to be honest, I think people overstate the changes. They did not last. Lots of salesmen in this industry like any other.

Best of luck to JPLers. I know someone who was laid off. These are some inordinately talented people.

I get the strong sense of the latter. It’s not actually the SEC that decides what’s illegal or not - these charges they speak of are for the judicial branch to decide on. I don’t follow crypto much, but it sounds to me like there isn’t a legal ruling precedent yet for crypto securities (hence Coinbase’s broadcast lack of certainty), so how could the SEC even give a guidance as to what’s illegal or not when they don’t know yet? But it’s definitely their duty to find out and establish that legal precedent. Given Coinbase is thinking about moving out of the country, I suspect they actually do think it’s not going to go in their favor, and might be trying to publicly blame it on the SEC versus admit they could have an illegal business model. Even in that case, I don’t understand the complaint because they got a good run - if the SEC had been faster and clearer up front, that party may never have happened.

Sleep well. Sleep is critical for structural learning (insight, generalization, etc.). Your hippocampus is theorized to randomly walk through memories to build these structural connections in your cortex during slow wave sleep.

It wouldn’t be hard to stop. Chip fabrication is a concentrated industry with a high barrier to entry (indeed there are only a few companies in the world producing high performance silicon using technology made by probably even fewer companies). Restrict AI chip making and the materials needed, and you’ll restrict AI. I can see global treaties between large nations effectively controlling production in the same way that we have nuclear treaties that prevent proliferation or enrichment.

Yeah this is bizarre. Imagine if a cop came up to you and told you to stop committing a crime. Being difficult and dismissing the assertion is going to land you in court. It’s your responsibility to know the law, and it’s certainly not the cop’s responsibility to spell out the public law for you before they can do their job. I don’t know what coinbase expected - actually I think they knew they had no case and are just delaying and trying to take the SEC down with them, given the attacks.

Wonder why this took so long. This was done by an influencer so no way it was under their radar for long?

Yep. I find it funny - though I feel bad for the small businesses that may have been oblivious - that a VC CEO is complaining when he effortlessly could have hired a financial consultant for a day to look into the first bank he encouraged his customers to use. A regional bank that had an obvious and alarming reputation of being the only place in town that would make high-risk loans to high-risk ventures. If anyone outside SVB was in the responsible position to avoid this situation, it was this CEO, and I’d suggest - if he truly believes in maintaining innovation and having those responsible suffering the consequences - that he personally go first in making his companies whole, rather than the taxpayer.

Are you suggesting tens of thousands of small business customers need to do due dilligence on the investment practices of their banks?

Honestly yes. All it takes is one financial analyst's time. I do it with my retirement plan for example, and I'm only a "small business" of one family. If there was demand for such info, I'm sure there would be a small community/industry for evaluating bank books like there is for financial planners (if that might not even be something a financial planner could already do).

And particularly with Y Combinator advising so many companies, I think it's on the side of negligence that they didn't evaluate the bank they were steering their companies towards. They were steering them there because they knew that tended to be the only bank that would deal with their high-risk companies - and it's too hard to believe that professional VCs didn't recognize that such a bank could have a lot of risk in some dark corner to compensate.

I understand your observation but I respectfully don't think the conclusion is the case - at a large company I have seen an internal report that shows a substantial issue with productivity that points directly to a lack of in-person interactions as the cause. The Amazon CEO not citing data in his decree does not mean there's no data, or that the data is poor, or that the decision is based on a gut feeling. Amazon is a self-declared data driven company with 1.5 million employees - they absolutely have run the numbers on as big of a decision as this and I would be surprised that a decision as high-stakes as this (both execs and WFHers would agree) would not have due-diligence data to back it up. It is just not shown. The CEO is exercising his authority instead. I think after 3 years of being unable to fix these issues, the decision is final and hence the reliance on authority as an argument instead of data. Giving people data to try to pick apart would just be an unproductive activity that would only serve to damage efforts - it's not meant to be a conversation.

I am confused by how this would avert some of the issues described in the article. It describes dustbowl conditions as evaporating water out of the ground. Does no-till drilling solve this?

This was exactly confirmed by an internal study at the company I work for. Engineers that depend on collaboration suffered a significant drop in output artifacts. Those being largely junior or mid-level engineers. Staff engineers actually did better, although ironically they tend to be the ones coming into the office consistently (anecdotally).

One of the most surprising takeaways was that while engineers in the office more often claimed they felt less productive in the office than at home, in-office engineers tended to have higher productivity and WFH engineers when the actual numbers were examined.

40:00 is the chandelier.

Salesforce? I read Benioff’s comments differently. He gave an example of lower productivity in sales, but that doesn’t mean it didn’t apply to engineering. It was just an eye-catching example that 50% of sales wasn’t accomplishing anything or whatever it was.

They are. My former employer briefly had WFH due to COVID. It ended with a small number of people getting fired and the rest coming back 5 days a week. People just weren’t working, and they had some astonishing IT stats that hinted at how bad it was.

Over at Google, there’s an internal study that shows raw code checkins have dropped dramatically since WFH. I’m pretty sure Sundar’s 20% more productivity comment came directly from the numbers in those slides.

It does not surprise me that Zuckerberg and Benioff have gone the length of calling out their employees’ productivity directly. I’m really surprised this comment is this far down, when it’s been painfully obvious at the couple companies I’ve worked at through the pandemic that productivity is the elephant in the room. Some CEOs dress it in corpspeak about collaboration, but if Google, Meta, and Salesforce are all complaining of worker productivity, I think that’s the true reason. I think we’re too afraid to take an honest look at how productive we are, which will ultimately be the demise of WFH we want to defend so badly.

Since this downward trend started around 2010, when I was in middle school and it seemed people where first getting on the internet en masse, I wonder if this trend is due to higher internet exposure in kids bringing widespread awareness to the financial reality of a humanities degree. It’s common to hear humanities degrees being derided on social media sites, but social media is also the only place I really hear these attitudes. I think this trend could be due to being more informed, as well as social pressure.

Thanks for the insight! Encouraging.

Somewhat unrelated but a remark of yours was interesting to me. You mentioned PhDs sometimes have previous engineering experience; did you find it common for ML PhDs to have waited a few years before entering a program? I’m exploring doing the same but had thought the ship had sailed, now that I’m in industry.

I figure most companies are too small for that to be budgeted. Though, it’s possibly a good selling point for cloud if it’s capable of it.

I’m not sure what these rooms control, but it’s possible the control room is simply part of a greater facility that has a mandate for this. Which makes sense, since there’s something to control here that may be more sensitive / dangerous than the control room itself.

It was a great experience (I was on the satellite side of development).

Hey, I worked on this. Glad to see it getting a lot of press :).