My wife had a miscarriage while we were staying at my Uncle's house in New Jersey. I was going to call 911, but instead, he decided he'd call from his landline. Difference between ambulance being free (because it was to a city resident's home) and costing what would have been $5k or so at least. Wild.
This sounds lovely, and ideal, but my read on this situation for, say, Google Drive, is a bit more nuanced: Anthropic's docs say that its permissions are scoped to the folders you grant access to the _user_ you create for Claude.
This creates an unfortunate situation where, for example, I can't scope a given channel's Claude Tag to _only_, say, a project's shared drive or specific folders, because I'm using a custom claude@ Google account in my org, and its permissions have to be all-encompassing.
It also means I can't escalate privilege and expose certain documents that I would like to: say I have a private #hr channel that I would like to have access to our HR shared drive docs with proprietary information. My understanding is that any user in any channel with @claude tag in it would be able to interrogate Claude about any file that the Claude Google user has access to, regardless of which channel they're in.
I'm trying to determine if the way around this is the alternate Google service account option, but that mentions domain-wide delegation in such a manner that it makes me think it would replicate the problem, and I obviously don't want to create custom Google accounts for each project scope in my org...
I have left comments on HN and Reddit and then found them rolled into Google's AI search summarizer OR the knowledge box with _extreme rapidity_ such that when I went, mere moments later, to go re-check some fact I had cited, I found my own comment at the top of the results, repeated back to me, but with authority and gravitas, ensconced in the austere trappings of the knowledge box.
I'm staying in a hotel right now and the TV is locked in hospitality mode and was blocking me from just installing Plex. It (Opus 4.8) gave me this whole jeremiad about how I need to be careful and it probably won't work and I should just watch on my laptop, but it did give me the service menu code. But man, it was such a downer.
Gemini gave it and clearly explained how best to get in, and then troubleshooted a few other weird issues that cropped up, without the moralizing.
Hard to put an economic damage value on the psychological scarring of everyone in the country sending their kid to school with this in the back of their minds, to say nothing of security theater put in place in an attempt to assuage those concerns. But, sure. Crowdstrike oopsie wiped out a lot more market cap, so I guess that's the priority.
The specific breadth of that oopsie, to recall, was that multiple human reviewers recommended escalation to law enforcement, and were rebuffed. So the system _almost_ worked except for an unforced error and people died as a direct result. Oopsie, indeed.
Because nothing about the IPO price has any resemblance to a fair market valuation, and if it's being propped up by this forced inclusion, even less so? The rules existed to fundamentally protect against a Potemkin village situation where an underwriter and some early round investors whip the valuation into a froth and raise against a rabid corps of retail investors who don't necessarily care about a PE ratio of 1,000+ because they're buying the hype.
More importantly, it allowed organic price discovery to occur. This eschews that process because the indexes are _forced_ to participate essentially at _any_ price, so rather than the market writ large having the opportunity to reward or punish the underwriter pricing of the IPO and determine any true idea of price, they're forced to buy the banker's narrative, which will intrinsically prop up the stock to some degree, but at what cost, and based on what underlying?
We're building something along these lines, but since our roots are a consulting business, we're still building around the idea that there needs to be an expert integrator doing the front-loading work of discovery/decomposition/scoring of tasks/implementing them as those agents. These tools are terrifying to anyone not quite technical, and it turns out, people are bad at decomposing their own work, let alone describing it in a box with a blinking cursor.
We're obviously going to be holding ourselves back in terms of scale and in terms of not being a "true" SaaS with this approach, but my thesis is that we get much higher quality results and higher compliance/activation and can charge more for the bespoke model backed by our own platform.
This is the craziest part. We've long had publicly subsidized private projects, or corporate tax breaks given to entities with the fig leaf explanation that the projects will bring economic activity, or jobs, or some sort of long lasting, durable benefit, or even at its most craven, keep a stupid populace happy that their favorite sportsball team hasn't left their town.
Datacenters do none of these things. They don't bring any true employment numbers to make a difference. They don't materially improve their surroundings. They don't increase land value. They aren't an attractive neighbor. They don't benefit the tax base, not least of all when you're doling out massive tax breaks and deferments... no one _wants_ them, so it's not even like this is a "well, you're just not a Buffalo Bills fan, but the rest of us are, get on board!" situation.
There is zero net benefit to the citizens in these places to welcome these facilities to their town. Shorter latency to their chatbot of choice, maybe? But in any practical, moral hazard sense, these are all pure net negatives for the communities, and it's wild that these leaders think they're some sort of marquee, glamorous, prestigious win of a project.
Compare this with other, things-that-rhyme-but-aren't-the-same projects, like the TSMC fabs in Phoenix: these projects are bringing a ton of high-paying, new jobs (and, somewhat controversially, an expat community from Taiwan to help onboard them in the meantime), but they're also delivering in other economic terms because of the supply chain's knock-on effects: the TSMC fabs further the reputation of Phoenix as the Silicon Desert that Intel, OnSemi, Microchip, and Motorola had long been working towards, but at a much more amplified scale, and in a truly meaningful capacity. The money being spent here is staying here, and driving some real practical benefits. But even still, it's an open conversation around how careful we need to be with the water usage of these fabs (though TSMC is aiming towards 90%+ recapture in the next few years, I think it's ~60% right now), and other considerations... they are still, on balance, bringing 6,000-12,000 direct jobs, and even more indirect jobs as they continue to expand.
These datacenter projects don't even do _that_ well. They're just upsetting the power grid and creating unfortunate microclimates for the immediate vicinity for a handful of NOC jobs. (And some itinerant construction and engineering jobs.)
This article in the WSJ was by an author who decided to induce immunity to poison oak by exposure: eating it. It's... quite the journey, evidently, but possible:
Right, but there's a core conceit we use in the US (mostly) that you are innocent until you are proven guilty, and if you are wrongfully accused (as was evidently the case from the author), you should perhaps NOT be put into such a grim set of living conditions with essentially no rights.
In this case, the author evidently _was_ a law abiding person, so the optimization failed, senselessly, likely out of a systemic effort to strike enough fear in the populace to over-index towards avoiding the possibility of this sort of situation. (Much like Singapore caning people for minor offenses.)
Whether or not you agree that such draconian punishments or processes are effective or fair is a different discussion, but this person was LITERALLY not supposed to be in jail, so how fair is it that they were removed from polite society for over a month in such poor conditions and at considerable expense?
They _do_ specifically protest, and it's crazy that they're able to detain you like this from an accusation while they build a case, even if you're innocent. In the US, barring flight risks and past history or cases of real malice or violence or an ongoing threat, you can at least typically make bail, AND the conditions in a jail are generally far better and less strict than this:
>Both cases were ultimately dropped and the second arrest was essentially tied to the first and shouldn’t have even been possible. But because of how the system works weather it’s a viable reason or not, they can still trap you in there for a time while the case is being reviewed. I met others who where there for shorter and much longer periods of time. The worst part was knowing i was innocent. After it’s all said and done you walk out and they act as if nothing happened. Not only was this was all extremely traumatizing but it cost me a HUGE of money that I really did not have and caused irreversible damage to my life.
I've actually thought of a much more dystopian idea: that Waymos could be technically used as roving traffic cameras, and report on the human drivers around them. They absolutely have strong enough telemetry systems to be able to determine things like excessive speeding, dangerous lane changes, red light running, etc., and their imaging systems could probably pick up a license plate with little additional modification... it's obviously not great from the perspective of general optics and morale, but it would surprise me if no one had floated basically WayNarc as a business model...
Any time I've loaded something into the trunk of a Waymo, it pre-emptively pops the trunk when I'm getting out _for_ me and reminds me to get my things from the trunk, so this is... surprising as a failure mode. Wondering if there was some issue with the latch/opening system, because it's definitely programmed to work the right way. (Or he tossed his stuff into the trunk from the main cabin, but... it's a pretty low hatch ceiling there.)