I don't understand why people think a lot of these things are 'products'. I have seen a ton of security industry stuff in the last month that once you peel back the marketing glossy, you find out that it is a just a tool that generates longer more verbose prompts from simpler ones, and for the better ones that means insertion of content from the 'system prompt' so you get something that is at least considered and somewhat consistent. The crappier examples just run with whatever the model tends to spit out in absense if inputs beside the prompt.

Next they shovel the four pages of instructions they wrote into claude code or codex where upon some MCP or skills that just wrap whatever APIs their product already had 3 years ago do all the heavy lifting, again these are the better ones... the crappier examples let the model dicker around with trying to use curl, and bash in some container to do unverified/validated operations which may or may not lead correct results even if they do happen to execute and return 0.

Finally some pretty charts with paragraphs of plausible sounding text underneath come out, but we are left to wonder if any of it can be taken seriously or relied on.

-- The shovelware isnt even shovelware anymore the who product is just fancy claude prompt. It is all getting downright scary, and painfully apparent a large portion of the decisions makes don't have the slight grasp on how any of this works.

Exactly this.

You don't have to look back to many decades to notice that older people were by and large in better health.

Survivor bias you say! Ok sure more people live longer now, but a lot of that has to do with injuries and acute illness being more survivable now than it was. What does the picture really look like when adjust for the people who did not die TB, Polio complications, serious infections, physical injury, etc. How does the chronic disease picture compare for the octogenarian cohort across decades then.

Which gets us back to red meat, alcohol use, cooking with tallow, smoked as means of preservation, salted as means of preservation, etc foods and lifestyles have been with us the entire time. it is a lot of virtuous things like eating our vegetables fresh out the garden not shipping in cold storage for days or weeks, physical exercise/labor, we have removed.

I have transacted a number of homes and land parcels, so while I am no insider to the real-estate biz I mostly understand the entities and relationships that exist at least in the non commercial space.

I can't understand why sellers would want this in the general case. Typically when you list a home or property you generally want to sell it as fast as possible. Narrow exception being you have not yet identified a new primary residence, even then most sellers will need the equity or be looking at bridge loan so you'd still be wanting to get the current listing under contract so you can understand what the cost and life time of the bridge loan are likely to be. Otherwise you are looking to get out because the long you own after you decide you no longer want to do, you're just being eaten by carry costs: fire insurance, liability (if it just a parcel), utilities, taxes, and maintenance. In summary you want to solicit the largest number of competing offers from the biggest pool potential buyers as quickly as possible.

You are already paying your seller's agent commission, they are supposed to be working for YOU, in what world again under the general case is hiding your listing from potential buyers even if only for a short time good for you?

Now imagine you are buyer, again you'll be paying commission, the agent should be working for you, in what world are better served if chose to limit the potential listings to match your with vs the whole of the market. Zilow is easy, if I was buying I'd scroll thru the listings on my phone and for sure if I saw something that looked good the agent hadn't matched me with I'd be ask why, and I'd want a decent explanation how come they did not include it..

I do realize there are unique properties which might you might want to market more selectively. A land parcel for example that is a good location but will require significant title and easement work before its accessible/buildable/farmable/whatever; sure you might shop that to only sophisticated buyers who understand the details, costs, and risks there because otherwise you're going to get a bunch of offers on paper with standard contingencies, that have you going - dude it is right here in the listing there is a utilities easement but the local power company will need their name added as user or have the easement made public; so already we don't have a contract can you read? You need a buyer who gets that they are going to have to probably bribe^H^H^H^H^H compensate the neighbors to add name to easement because they might figure it is nicer if nobody builds anything there, on the other hand they might consider the improvement to the value of their own property should the question of getting power if desired be removed, but someone has to do the organizing work of getting all the parties to act.

Similarly buildings that are 'historic' and such again you might want to qualify the buyers.

People with my genetics start dying around age fifty - polycystic kidney disease.

I'm ten years older than that, my blood pressure earlier this evening was 120/70something, my last fasting sugar was around 85, I weigh ten pounds more than I did in college, and I've walked 1,393 miles in the last year. My last vice is caffeine and I will go off it periodically, in one instance for seven years.

I got dealt a terrible hand, health wise, Lyme at forty that triggered a complex immune condition, but I refuse to feel (or look) bad. A lot of it IS in your hands, you just need the will to change. It's not easy ...

Nope. But in my experience, in well-designed cities, you rarely need to walk more than about 250 m to 500 m on either end of the trip, and if you're working in a dense urban core, you often get delivered right to your building or very close to it.

Ah, yes. Another "one day people like me will force people like you to live the way we think you should" post.

Keep dreaming!

Personally, I want a self-driving RV. Go to sleep and wake up at the destination.

Yes, the magic house.

Twenty years from now the techno-nomadic lifestyle will be commonplace. Why own a home that stays in one spot when you can own a home that lets you wake up in a different destination whenever you want? Think of the modern RV culture, but multiplied 100-fold.

A Waymo taxi is probably around 4.5m long and can hold 5 people. So 28m worth of Waymo taxis bumper-to-bumper with zero clearance between them can hold 30 people... less than one-quarter of the tram. In reality, you're probably only going to get 15-20 people in the Waymos because of the clearance between them.

Individual vehicles are just about the worst way to move a lot of people efficiently.

And will that tram pick up and deliver each of those 130 people directly at the doors of their homes or workplaces? Because I have yet to see a bus, trolley, or subway car that will do that for 99.99% of its passengers.

Waymo can do exactly that. That is exactly the reason why Waymo and Uber and Lyft (and taxis!) exist. Mass transit cannot solve the last-mile problem.

In Arizona, the police routinely have to rescue people who drive around roadblocks into flooded arroyos and wind up with their vehicles floating down the wash. It got so bad while I lived there that the city had to start charging people a fee for their rescue to curb the stupidity.

A decade from now, I have no doubt that the authorities will still have to rescue drivers in flooded roadways. I am also certain that Waymo vehicles will have stopped making that mistake years earlier. We can fix autonomous vehicles. We can't fix humans. I'll take the Waymo, thank you.

What is the different problem that they solve?

Here's one: the Waymo will never sexually harass a female passenger. And yes, a lot of women choose Waymo over Uber, Lyft, a taxi, or even public transit for exactly that reason.

It's amusing to see people assuming the "Get off my lawn!" roles of their elders over AI and autonomous vehicles. Not so long ago I recall similar "shout at the sky" attitudes on Slashdot about Uber and Lyft, which have now become the "good guys" because they employ human drivers (who sometimes sexually harass their passengers, of course).

Waymo is giving half a million rides a week right now, and that number should double by the end of 2026. Even if you could somehow shut down Waymo tomorrow, a Chinese company would move right in and take its place, because the technology is not going away.

The issue is Microsoft has their own people who have mission and authority similar to CISA but scoped to the organization as do many of the other institutions I see making this class of error.

It isn't that institutionally they don't know better, or even individually they don't know better, it is an operationalization problem and there simply exists to much pressure in terms of time time to be sloppy with a credential, coupled with the near certain knowledge that even if that sloppiness is process or policy violation it is sure to go unnoticed or at least unpunished unless something bad happens and even then it still might not carry much in the way of personal consequences.

Fundamentally SaaS/PaaS/Cloud security is far to reliant on not just everyone knowing what they are supposed to do, but actually doing it dependably and consistently everyitme. It simply does not work at scale.

Zero-trust just isn't a very good model over all because it makes everything about identity and discretionary access management, and people are just not that good at identity management. They are better about DAC, but even then there is a lot of templatation to just say sure give'em repo access.

That is going to be the practical result here.

Some code will get released, it will be most vanilla foss projects + a driver or two. You might even be able to build it but your won't be able to sign or run it. Version next will ship some generic kernel module, that provides some ioctl hooks or something and they'll move the drivers into use space, so they don't have to share those either.

Maybe if consumers are lucky there will be some groups of discontinued models where thanks to some signature checking flaw it is possible to monkey with the software without destructive or likely to be destructive hardware modining, and you'll have a scene like when people were running around hunting for v1 - wrt54G accesspoints.. for a some years, but as a practical matter not very many people will get anything valuable.

TIVO-isation is a problem manufactures have pretty well solved. I have to say Linus's unwillingness to try to migrate mainline Linux to GPL-3 has really hurt consumers. It was probably the one project with enough technical weight to have forced some hands, but it also probably would mean a lot less Linux out there today as well.

maybe partly, but the reality I know as someone who reads a lot of penetration testing reports, is big supposedly mature organizations end up putting useful credentials (as in not just some QA mock enviornment nobody cares about in CI/CD stuff) in their git commits, all the freaking time.

Cloud security is a s*** show a lot of places, even places with mostly capable people, it only takes one idiot or one careless person to really mess things up badly. That is the problem with PaaS/SaaS model generally.

if it's so easy to control, why on did a trained nurse in a US infectious care facility contract it? Never underestimate the ability of governmental incompetency when it comes to letting something happen that shouldn't happen

https://www.nationalnursesunit...

Except it absolutely is Embrace-Extend...

It is embrace Linux, just so long as your are running it on their compute...It is extend Linux,they have already used their influence to stuff all manor of rather cloud-specific tooling into systemd, and successfully crammed that stack down on the broader community.

Finally it is extinguish in the software freedom sense the GNU side of GNU/Linux always cared about. Unless your are like beyond careful about every component you use, every bit of tooling you chose, and every other architecture decision you make the odds of anyone not a large enterprise being able to shift their application from Azure to some other cloud or their own compute/hosting is low. If you really do go the truly cloud agnostic route, you'll be giving up a lot of the value add features of the platform and paying a higher bill at the same time.

I don't disagree but the problem is identifying when we jumped the shark. Which I think happened damn near 100 years ago now when the SCOTUS blinked and let a lot of the New Deal happen.

We have now gone so long without 'keeping up with Amendments' in terms of actually enabling the Federal government to do so much of what American's of all political stripes currently view as good/necessary/appropriate it is really difficult to take "Textualism" to the logical destination it really ought to be taken. The 'soft textualism' we get from current conservative wing of the court is probably the best we can really hope for.

If you really for example did a legitimate read of the 9th and 10th amendments, probably half or more of Federal laws are unconstitutional or at least could not be applied to 90% of the instances they are. It would break our society...

This is real problem people who immediately shut down conversations around national divorce, or moving toward greater State level sovereignty as in letting people start thinking of themselves more as "Virginians, Floridians, New Yorkers", etc rather than "Americans" can't accept. We let a 80 years of sloth and neglect pass by as far rigorously applying the Constitution and using the Amendment process for real rather than feel good issues like Senate elections, and as far as keeping the American experiment on course, I am not sure you can get there from here now.