smartician

I see. It sounds to me then this is a knee-jerk reaction due to the app having been vibe-coded, not so much due to how the app works.

What sort of "grief" did you catch specifically? What are people's concerns with your system? Privacy, resource usage, etc.? Like you said, of course your thing uses AI, just a different flavor than what's currently en vogue.

If they're a PaaS, isn't it insane for them to run their own infrastructure and the customer workloads under the same cloud account?

Multi-region failover?

Full disclosure: I'm a Google employee, but not in the areas mentioned in the article. This is my personal opinion.

Regarding the Nest thing: I don't think those devices stop working completely if you don't enable location sharing for the "home and away" feature. It might be bad UI that made the user think that this is the case?

Regarding photo sharing: I think that permission is necessary to show a "photo picker" inside the app that allows the user to pick and choose which photos to upload. I'm not quite sure what the alternative would look like: "he can identify specific pictures in his library and grant access to just those" --> How exactly would that work without the app having access to the pictures? Also, does the author believe the app would then secretly analyze all pictures and send content back to the mothership without the user's consent? Again, this might be a communication/UX issue...

If the goal is to emulate human behavior, I'd say there is a case to be made to build for the same interface, and not rely on separate APIs that may or may not reflect the same information as a user sees.

Research seems to suggest we need exponential training data volume increases to see meaningful performance gains: https://arxiv.org/abs/2404.04125

Personally I think we've already hit a ceiling.

Define "AI Agent"?

IIUC the EU basically had to force banks to implement this via a mandate. I don't see "business friendly" US lawmakers doing the same anytime soon. The joys of the Free market™!

There's an interesting podcast about this: https://www.youtube.com/watch?v=XpkNGMSWbYk

They have three months to find new roles/teams. Their employment only ends if they can't.

My completely unfounded take: Culture.

In Europe, you have a millennia-long culture of classism, feudalism, aristocracy, guilds, etc., that's still somewhat ingrained in modern society. People are more content with being born into a certain class, not expecting or striving for upward mobility. You'll see people picking a profession after school and sticking with it until retirement, and it's out of the norm to switch or try to start a business.

Whereas in the US, a few hundred years ago settlers came in and had to start from scratch. It doesn't matter what you were in your previous life, everyone had an equal start (more or less), and they had to build whole cities from scratch.

I think this partly also explains the differences in salary between the US and Europe. There is a deeply ingrained sense of "ordinary workers shouldn't make more than $X, that doesn't make sense" creating a kind of tacit collusion between employers, which does not exist in this form in the US.

Did you previously search for that attraction either on Maps or Google Search?

I don't know who Matt is, but I now know that you lack even basic financial knowledge. According to your logic, if I can rent a house for $1, then I'm still losing $1 and my landlord is "gaining $1 in property value". Looks like it was a bad idea to insert myself into this conversation in the first place.

You're assuming that all landlords are rational. Many either see it as a leveraged investment where rent is just one factor to offset their cost, hoping that appreciation will make it profitable. Or the home had strong appreciation so that their cost relative to the current value is low. Or they have some sentimental attachment to the property and rather rent it out at a loss than sell it. In California especially, they might have locked in a low property tax rate which is a fraction of what you would pay when you buy the home today. For example, my in-laws pay $700/month in property tax. If I bought their house at market price today, my property tax bill would be $3,000/month.

For example I rent a home that's currently worth $1.3 million for $3,500/month. The landlord bought it in 2004 for $570,000. Would it make sense for the landlord to sell it instead of renting it out? Maybe, but based on their original investment they're getting a good return. If I were to buy that home off of them, my monthly payment would be more than double the rent. So this works for both of us I guess.

I'm an SRE.

* How many days per year are you on call? About 50. Depends on whether the team is fully staffed.

* How often do you actually get a call? Right now I think it's about 0.3 pages per 12 hour shift. So once every three shifts.

* How many rotations are you in, and how many other people are in those rotations? One on-call rotation with ~8 other people. Fluctuates wildly because SRE is a high turnover job, but right now we're pretty much fully staffed.

* Do you get any time off or additional compensation for being on call? Two thirds for time outside business hours. So a 12 hour shift on the weekend gives a full day off (or the monetary equivalent), even when you don't get paged at all.

* Do you get any time off or additional compensation for responding to an incident? No, except for a potential spot bonus for handling a huge outage.

I grew up in Germany and worked there for eight years, now I'm in California for 15 years. Now that I had children, I'm contemplating moving back, possibly to Zurich, which would allow me to maintain a comparable level of income.

The main thing that's holding me back is the support my wife's parents who live close by are giving us. Without them, navigating the daycare landscape is difficult and extremely expensive. So that would be a huge plus for Germany in your case. In terms of career, California has the obvious advantage, although that advantage is not as strong in Sacramento and San Diego as it would be in the SF Bay Area.

So it all depends on what you value and what your gut feeling says. How do you like Amsterdam?

Another thing to consider: If you have two options and both have strong pros so that it's difficult to decide between them, then either choice will be fine.

So he thinks one of the Googlers he worked with convinced his employer to fire him, and now he's suing Google? Shouldn't he be suing his employer?

https://medium.com/who-is-the-american-hindu/why-do-we-say-n... is one example (this is the refutation, the actual article by the lecturer is linked there). Another is a lecture titled "Why Savarnas rape", as a reaction to the series of rapes in Uttar Pradesh in 2020. Savarnas are basically the opposite of Dalit, namely members of the upper caste.

I find it very interesting that the group in power is trying very hard to suppress any criticism, basically silencing the disadvantaged minority (within the Indian diaspora).

I just found out that my new phone can't pair to my 10 year old car anymore. I googled it and it may or may not be a new version of the protocol that's incompatible with my car. I might have to go see whether my car has an AUX input, since I never take calls in the car anyways...

A sample of recent departures:

* Notion * Truepill * Plaid * Cresta AI * Meta * Plaid * Meta * Roblox * Roblox * Amazon * Spotify * Chainguard

This is just a sample of those who left in April and publicly shared their plans. Also lots of "taking a long break", "going sailing", "retired", "traveling", "riding bike across the continent", etc.

You can log out of all, then only sign in to the account you want (and optionally remove the "unwanted" account(s) from the sign in list).

Multiple things could be at play here. Google Search runs hundreds of experiments. Every time a parameter is tuned, or a snippet of code in one of the dozens of components that make up search is changed, that change is being canaried in an experiment. Users are randomly assigned to cohorts for these experiments (either based on their user ID when they're logged in, or a temporary user ID when they're not). Also, every search request will hit different servers holding different states, and sometimes processing components may time out intermittently. As a result, every user will see a slightly different set of results, or experience other diverging behavior like the truncation you're seeing. This might give the impression of personalized results, which in reality is not happening (only ads and other content that enhances the SERP are personalized).

An SUV, a URL, an RSU, a YMCA... It depends on how you pronounce it, not how it's spelled. Common pitfall even for native speakers.

Partly tongue-in-cheek. These review processes exist, but whether they're required or not depends on the product area and type of project.

It's not much different today. Nowadays you'll also need privacy review, accessibility review, security review, and diversity & inclusion review.

I don't have to return to the office until January (for now), but am voluntarily going twice a week now, just to take advantage of all the amenities, bond with co-workers and have a change of scenery. I think I will be happy with 3x per week.

I could apply for full remote and would probably be approved.

What's ENG?

I don't quite understand how a settlement in a bankruptcy case can include an immunity provision against future criminal charges. What is the mechanism for that?

Even though I'm not a blockchain expert, I'm a bit surprised by the other responses.

AFAIK, when a node finds a new block, it announces that to its peers and starts working on finding the next block. No explicit acknowledgment from peers is required, instead they implicitly acknowledge the new block by appending it to their version of the chain and working on finding the next block based off that version. If a node receives a broadcast with a chain that is different than the one it currently works off of, it keeps the one that is more difficult to verify ("longer"), discarding the other one.

If you would divide all nodes into two unconnected sets (by the hypothetical firewall), two parallel versions of the chain would emerge, and those chains would diverge at the block that was last found before the separation event. Once you drop the firewall, the longer version would "win" over the shorter version, and all blocks mined on the losing chain since the separation become "stale", and the transactions contained in those blocks are returned to the mempool for re-processing. This is a huge problem, since the processing order is not guaranteed and depends on the fees transactions are endorsed with. Addresses could theoretically spend BTC they do not yet possess. Not sure how this would be handled.

If this is a prolonged event, the Bitcoin community would probably work on a software fork in some way?