This is worth a mod up ^

5k is a heck of a lot of headcount for what Mongo seems to have in terms of scope of product offerings and support services.

At minimum people have a duty of self care. Why are they expecting others to do that job for them!

I am not even arguing we should not help in hard cases, from both a policy and an ethical standpoint. However suggesting the in-action someone makes us responsible for those outcomes is insane.

Every type of law enforcement ever, has killed an innocent or two; especially in situations where people are protesting the law and large numbers of people are propagandized into obstructing law enforcement. ICE has a pretty damn good record here actually all things considered.

If you want to assign responsibility for these deaths, maybe take a hard look at the policy, organizations, media, legal system, prior executives who created a situation where we had massive numbers of people in unresolved extra legal status and allowed that problem to continue compounding for three decades. That is not party thing, either plenty of inaction under Bush etc to point at. Here is the dirty little secret about the who immigration conversation. Self interested liars who have a lot of media influence keep claiming Trump is the on violating the law here, when in fact he is the first person in the last 30 years to actually try to do what the IIRIRA actually requires the federal government to do. The Constitution requires the president to " âoetake care that the laws be faithfully executed", literally Clinton, Bush, and Obama, should constitutionally speaking have faced impeachment and conviction over IIRIRA!

No it hasn't killed a single person, not one. The conditions the locals have allowed to persist in places where USAID formerly was providing whatever...it was providing is what killed those people. Inaction in the form of lack of prevention where it requires concentrated effort and importantly you are not the proximate cause whatever the calamity might be in the first place is not equivalent to acting (killing).

Where was the EU? Where was the CCP? Where was the ANC? Where was OPEC or for that matter Walmart Corporation :-P How come none of them stepped in to fill the gap. What absolves them of culpability?

Personal shoppers etc are a thing; but what is different here is the scale.

If Amazon's experience is so bad, that is what that is maybe the market will move to walmart.com; not really the point.

The point is that letting the AI Boys become an intermediary between every enterprise and most people will NOT be a good thing. I am not saying Amazon isnt a bad actor or isnt hypocritical here on a broader context. I am saying that shop keeps of all stripes should be at least free to post a "No Robots Allowed Sign" and have it respected.

Because I don't think we will much care for the outcomes when we have a world where Coca Cola and slip Meta half a a couple hundred million to make sure the next model generally chooses their products when someone asks the AI to ensure the fridge has colas in it. The consequences for the market where there are six or so middle men everywhere won't remain 'free' long.

At least the retail arm of Amazon is um a wait for it a retailing operation. The individual candy bars, and $2 for 20z chilled sodas are up by the checkout line where you are standing around waiting at the grocery for a reason.

At any large retailer (an some small ones) quite a lot of thought goes into what is on the end caps and what gets less prime real-estate. I actually don't think it is unreasonable Amazon should seek to control the shopping experience to a degree even if to the point of using so called 'dark patterns' to drive some impulse purchases or steer customers toward higher margin items in the category.

That is consumer selling and has been since probably some ancient farmer bartering with his neighbor suggested he'd take fewer animal skins if the other guy agreed to accept some of the squash that had already gone a bit soft over the fresher merchandise.

Does big tech, take that probably way to far - I think the answer is likely yes, but how far that is and how they should be restrained is really a different conversation.

The separate question is should Amazon have to let Meta (or anyone else) step in and completely separate Amazon's own customers from interacting with the store front Amazon wants to present to them? I think the answer to that really should be no.

The reason it should be no is not to protect Amazon but because as much influence as they have in the retail space (maybe to much) think how much influence a handful of frontier model AI providers might get in terms of determining product selection etc if shopping via AI bot becomes the nominal consumer expectation.

One man's "gate-keeping" was everyone else's curating. The old model printing and distribution was a very real cost, usually two large to be borne by an author and certainly by a new author. The primary sales channels also had a premium on space, you can only put some many books on shelves low enough for customers to browse, their are only so many end aisles privilege products with placement showing more than their spine, and the story with magazine racks wasn't much different.

So the publishes provided enormous value to retail by both ensuring quality control, and filtering for what the public wanted to buy or was worth convicing the public they wanted to by thru various literary reviews, direct advertising, etc.

Digital has removed the *need* (and many of the retailers with that).

The question is was and remains how much value was the curating, QA to the consumer. We were reading/listening to better content because of the filter and not wasting on time on drek, alternatively what price were we paying for the gate-keeping aspect in terms of not hearing voices and ideas that might have proven interesting. What is the cost of the fan out. When everyone was pulling from the selection on the shelf at the local Books-A-Million, Dalton, or library shelf, there was good chance of running into someone who'd read the same thing you did. Now - FAT chance if someone actually read something it was probably a substack, or some Kindle E-book by someone you never heard of.. As with water-cooler television even if what we are consuming is very good the opportunity to have any local discourse about or discourse within your own social circles whatever form they take is nil because you must first convince someone else to make the investment in it.

It certainly is having an effect on social discourse and how ideas advance and change in our society and I think very likely calcify because their is no collective train of thought. As to the topic at and AI is just going to pour gasoline on all of this, it will absolutely eat the proofing, editing, polishing value the traditional publisher can still add in the ebook era. The incentive for various retail channels to 'stock' the stuff of least providence will be huge, if an algorithm can see sell Christine Middle Schooler another book about sparkling vampires and her Dad another story about manning a first generation aircraft carrier in the Pacific, she likes those and so does he, and they can pick from ones with almost no royalty/residual attached, more profit for them. If AI can make those good enough to get them to come back for more; what need for a traditional publishing house is there, that would pay the bills anyway?

for once I am really rooting for the extinguish part of Microsoft's Embrace Extend Extinguish MO..

The language and the tech are fine but the proponents are mostly insufferable.

I'll bet; my first reaction to that would be "Today is not going to be a good day"

Here let me google for you

https://www.sciencedirect.com/...

I hope it is one of those fancy engineered septic systems, with pumps, timers, floats, dose controls etc. In which case yeah ... I often wonder if the engineers and the health departments understand them either. AI is probably great diagnostic tool there.

On the other hand the conventional ones really are just a hole in the ground for the most part; if you need AI for that we are in real trouble.

I think it is more simple then that. It is about protecting the status quo.

Nobody is really happy with US education outcomes. It is painfully obvious that places with the NEA/AFT enjoy the most influence, is highly aligned with the poorest performance.

You have similar orthodoxies in the higher-ed pedagogical programs where application of current state-of-the-art classroom and instruction theory continue result in only a long year over year decline in student performance..

So how do these entrenched groups remain so and prevent a revolt from parents and students, well one way to wave some flashy new tech or radical re-conceptualization of what being educated will mean in front of them like (you don't need to do math, you need to be able to code); then plead you just need to give us another decade and you'll see how this current crop of students turns out to be the most polished ever!

When it does not happen - repeat.

Because who would not be trying to parlay their association with a multi-billion dollar company into a next career step or second career?

All of those people are or were drawing salaries that would make most of us blush. They want to continue doing that. The reality is they are not adding value and their know it. They are pontificating about what machines might do with large matrices of numbers, tokenizaiton, attention algorithms, and fancy looping constructs over it all. You can get that free on Slashdot!

fundamentally their options are -
1. Snow some other organization, commercial, political, etc into believing they were not being taken seriously but are really necessary and they should hire and pay them for version 15 of their safety whitepaper informed mostly by 1960s sci-fi tinged with 1990s political correctness.

2. Convince the public the world is going to end, like every cult leader before them; so they can do the paid speaking circuit or write some sci-fi novels of their own leverging the fact their names are already in print to rise above the noise and get past the mail room at McMillan.

3. Get a real job actually making something for 1/4 the pay, or less.

It is not hard to see hard that even if these are very smart, very successful people by any measure. Probably smarter than me, they are still "failing upward." If they were really all that they'd have been on the teams building the things, not occupying some position created entirely for the sake projecting an image of corporate responsibility. I doubt Durrant had anyone at Union Pacific churning out copy on the potential loss of cultural distinctiveness that might result if a Transcontinental rail line is completed, but he would have if he'd faced a public that had become preoccupied as we are today with the latest jobs report. He also would have found a reason to dismiss, drive them quit, or otherwise sideline them the INSTANT their recommendations stood in the way of the actual business activity as he saw it; just exactly as Si Valley has done.

You are right that problem of sandboxing an agentic system will still giving it access to outside information even by proxy is actually hard.

What bothers me is; these are smart guys who should understand that. They are building a testing these things at scale. I would *think* they would be very interested in questions like "hmm why does the system keep making the same http request over and over..."

  Existing SIEM-like tools should be spotting and flagging that top level behavior all day long. That ought to trigger various looks in the to the agent-to-agent message layers, and agent-model-loop. While the way they press releases read nobody every looked at it except from the agent message direction. Which is just beyond credibility. In the case of some the 'hacking events' I really can't imagine ordinary network monitoring did not spot some of the very noisy side-channels and strange seemingly off target-off task activity early. I remain of the opinion i was ignored and almost had to be if not quite deliberately so...done with a level of intentional ignorance, kinda like those mid 2000s VPN services that 'kept no logs' because 'your privacy' but we all know they knew a not in significant part of their customer base was using it for abusive activities like sweeping for open SMPT relays to SPAM thru..

There simply had to have been a lot of fingers jabbed in ear canals, and signing 'la lal la' at OpenAI and Anthropic over the past couple years.

This is one of the things I found I really don't like working with recent qwen models.

If you read thru the reasoning. It will write stuff like, "I was asked to ... but that will prevent me from ... the user probably did not really mean ... so I will ..."

I told you not to do anything other than read files under /usr/doc and search the web while trying to answer my question. That fact you can get around writing in 'plan mode' by running giant blobs a python one-liners, is not a plus.

These things have almost arrived at a state, I'd call "malicious compliance as a service." Which I do believe to be a training problem not a fundamental technology problem.

Ultimately it won't matter either way. Open source models will get better. They don't even need to achieve absolute parity with the frontier models they just need to get to 'good enough for most tasks most normies need to do'.

Some of them are probably already knocking on the door of that, and where they fail the frontier models often do as well.

Even if GenAI has dramatically increasing your business productivity, $100 a month per white collar workers becomes a pretty big expense item on the balance sheet pretty quickly. GPUs get better and cheaper all the time, eventually the market will deliver memory at reasonable costs again.. Hardware under the desk is usually cheaper than hardware behind the glass or in the cloud. Especially as the upgrade cycle has stretched from 2 years to 5 and longer. The open models get better, the GPUs get a few more compute units, in some not so distant future 32 of VRAM becomes such that it fits in $`000 price target Graphics/AI Accelerator package...even if that turns the $300 mini PCs you are putting in the cubes into $1300 dollar mini PCs, well if you get a least a year of use out them its cheaper than general Claude subscription.

Given the cloud AI boys can't make money at that price point now, it should be clear the model of selling everyone and the brother a subscription isnt realistic. It is going to make more sense for most people to 'own something'.

This is not say if you hosting some server with agents, or doing very specialized engineer or medical work as as opposed to more general office work Cloud AI won't be the sensible path forward. It very well could be, probably will be. If on the other hand your job is cleaning up the South East Region P&L spreedsheet for the your chain of HVAC dealers probably not so much; or if you're a home maker need to work up Dinosar birthday invites for little Tommy's sixth..