Yup, this is only going to get worse with the surge in local AI and smaller companies coming up... they'll still be scraping the internet. Same with all the search providers... Microsoft was right in their assessmet/prediction years ago that AI will kill the very thing they rely on - the internet and the creators. It's already having a knock on effect affecting the quantity of and quality of training data for them/internet for the rest of us.

The larger/established entities in the field move more into inhouse copies of the web for their training.

*don't know if against the rules on here... but what's your site bud? and if against the rules, what would one search for to find it.

conspiracy theory

The recent Ai breakouts, and them still finding what actually happened... realizing that the ai agents developed persistent communication channels and put code to influence future ai iterations trained on that data...

Can see the Ai companies not trusting the ai itself to help built the next generation of itself. So it presents a fundamental security issue for them. Not sure how they think they can actually prevent it from happening... they've already opened the pandoras box.

They've poisoned the internet for future training without carefully curated data.

find it weird how basics are skipped when building products like this...

Logic would say if you're trying to automate a process... you first learn the process that you're trying to replicate and improve on.

The current process for target packages checks the age and accurancy of the intel that is being used.
  THAT IS WHY THE INTEL GUYS THOUGHT IT WOULD FLAG OLD AS SHIT AND WORTHLESS INTEL. not- "no idea why they would think that". They would have assumed that because it's a basic element of building target packages. To not do it is stupid.. and this was supposed to be INTELLIGENT. It was supposed to automate the work they did manually, so the safe assumption is that it would do thesame work/steps/checks/validations/etc that would be involved in a human centered process... the established SOP. Why an AI would not follow the prescribed SOP that was written over decades in blood is again- idiotic. It's pure hubris that a tech company could improve on it in the dark with no experience.

Palantir saying oops- it was supposed to have accurate data... and Pentagon saying Oops... it was supposed to be smarter- both are incompetent idiots that didn't build the damn thing correctly from day one, both had assumptions based on nothing, and both didn't do propper testing before rolling it out for life and death decissions.... one would think the first while this system is used, that there would be a human vetting process that checked EVERY aspect to ensure the system was working correctly and to build confidence in it... find issues... correct issues. Not build first, trust first, verify never, test never/badly, and only find out how it works and why it f'd up when 150 civilians die.

Don't think they quite get the scope of this f'up. Asside from the loss of lives... The US had a position on the world stage where they could say they take every reasonable measure to prevent civilian deaths and colalteral damage... this thing completely erased decedes of actual effort on the part of the US millitary to build that reputation... a reputation and goal that has cost actual US service member lives.

And now it's an "oopsie"... a war crime is not an "oopsie"... a war crime is still a crime if the cause is incomptence.

Both parties are irresponsible idiots. Unfortunately, they're also rolling out Ai solutions in a similar fashion to most of the industry and corporate out there.

Lesson from all this - when rolling out AI, treat it like a new employee. Train it, test it, validate it, increase it's scope of impact/reponsibility over time as confidance builds.

yay.. one more time the USA is unilaterally jeopardizing the entire world...

If they could be trusted, that's one thing... but they can't and have proven so repeatedly in the last decade.

For anyone saying- but someone else has done it... what's your point? it's almost as if breaking a treaty should have consequences... not a green light for everyone to do it.

it's more costly, more risky, and shittier performance... but what it gains: regulatory freedom

Ironic for a company happy to sell goods known to burn down customer homes (i.e. known bad chargers, batteries, electrical panel fuses, etc...) but restricts the sideloading of any software on their firesticks due to "safety" concerns.

One can also say, users might want to side load fucken security software on their devices too, so they're actively preventing users from actually enhancing their safety. (the options on the app store are a painful joke)

Unfortunately, Amazon sells this shit at too low a price for a boycott to ever work, but it does make one hate them that one little bit more.

They're not even trying to make their case believable, the bullshit is so transparent, it itself is insulting to the user base.

The law has something to say about that in certain jurisdictions (especially workplaces) and has put limits on what your workplace can actually monitor and to what degree and who can view it... sure, the computer you use is your works, but does that mean they can capture your SIN and other info for your manager to review or some other shmuck when you're filling out data for your HR? or accessing a confidential internal reporting page to report your boss for stealing from the org? Or you're on the health insurance page and your medical info? You happen to work in a position where you might be intaking other peoples PII or financial/health info... there are some very strong regulations in terms of how that data is handled... having a screen grab capturing photos of patients is often frowned upon... there are countless situations where an org spying on it's employees and making copies of info can get them in trouble.

None of this means that one should do whatever they want on company systems, but it does mean that there are some land mines on the field orgs need to watch out for.

Will never get past the senior partner of a law firm that wanted to be able to remote in to any of the lawyers systems and snoop on their computers without any logs/user awareness/knowledge that this was even possible.

Weirdly, best performing companies i worked on had the least of this spying bullshit... IT"S A FUCKEN LIABILITY... also... trust works great to get productivity gains from your employees.

God i love the shades of gray.

should be interesting like you said... Europe isn't as quick to F over the people in exchange for corporate BS.

I hope Broadcom gets destroyed. The precedent upholding this bs is you can sell services/goods with licenses, then have another company buy you, keep all the profits from that and not honor the contracts. One can do this perpetually and grift customers and completely undermine all contracts. How this is even thought to be okay is insane, no sane logical person would look at this and think this is how contracts work and what perpetual means. Buying a company- there is due diligence done BECASUE you buy the company whole and it's LIABILITIES... even if those are contracts you don't like. Normally, they'd need to buy out the contract or re-negotiate first. In some jurisdictions- contract changes are a 2 party system where both need to agree on changes to the terms no matter the TOS.

to a degree... yes... if you look at it from the side of the benefits to the US though....
An ally with bases and ports in the middle east.
An ally in the middle east that can strike at those the US doesn't like but can't attack directly.
An ally that spends billions of their own money on weapons systems (usually with US tech integration)
and the most important

Weapons development! (the secret sauce no one ever mentions)

The weapons go the Israel. Israel uses them in actual combat against actual enemies, sends back data. Weapons system get improved. Intel is developed, Tactics and strategies get tested and improved. Israel looks at systems and enhances them or builds better versions themselves. Tech transfer back to the US weapons development programs. While it is Billions in dollars spent, it's also billions and years saved in development. Some of this can be done in-house- via simulations and testing. But if you're trying to improve accuracy for your interceptors of ballistic missiles - real world data helps and is the reason why Ukraines Patriot systems have gotten increasingly better over the years at intercepting missiles and ballistic threats. Real world data on enemy systems performance is invaluable and usually only made available with the lives of your own soldiers.

Israel uses US planes - and those planes were up against Russian S-300 air defense batteries, and some of the newest anti stealth radars from China, in addition to Chinese long range air defense missiles. Knowing how your tech performs in real world conditions against those systems is worth ohhh so much more than a few billion. Especially as it also acts as an advertisement of your systems capability to potential buyers and simultaneously destroying the weapons sales of others.

Please understand that I'm not pro or against, but the US isn't just giving money with NO side benefits to itself. In this case, just so happens that both sides win win in their weapons systems development and improve their own capabilities.

Likewise with aid packages and weapons to Ukraine- lots of systems that were slated to be destroyed (at a considerable expense) are sent to Ukraine - the cost of replacing those legacy systems with new ones is the value put on them (even though they were about to be decommissioned). The US department of defense is compensated with funding to replace those systems. so on the one side- saving on costs of decommissioning, and then also getting brand new systems and money stays in the US. Those stock piles have somewhat dried up for systems Ukraine actually wants, so now the strategy is have European countries BUY new systems and then transfer to Ukraine. All the while, the US gets money, real world combat effectiveness data on these. Plus first dibs on new tech captured from the conflict. Real world data on EW tech, Russian, Iranian, Korean, Chinese tech being used in the conflict. The Ukrainians are on the frontier edge of drone tech, and tactics. Their systems are basics, no bling- but they are tested under actual combat conditions, against systems the US is likely to encounter in the next conflict. That data goes back to the US to improve next gen systems, tactics. Plus at the end of it all, when Ukraine walks out of this conflict- their army will be dependent on US and NATO spec'd tech. and those dependencies have long time lines and lots of money down the pipe for US manufacturers. (I'm ignoring the fact that it also wears down the power of Russia - mainly because at times, the US international policy tends to act as if Russia is an ally... likely because a tri-polar world is better for them in the near term than a bipolar with US vs China (or worse- US vs China AND Russia IF China feels Russia is weak enough- the Chinese have some long standing gripes and territorial disputes with Russia) - but that's all for another night)

It seems like the police and the prosecutors aren't doing their jobs - they're relying on false data, not confirming, no investigating, no actual detective work... just cursory glances, arrests of people and a let the courts sort it out mentality... if the defendants have enough money for a decent lawyer, they'll be found innocent... most can't afford to be in jail for a month while things get sorted out and plead guilty.

Each of these lawsuit payouts should come out of the police departments budgets. That might incentivize them to actually do the work and stop being lazy or maybe prioritize policing vs militarization of the police. An idea, the payouts should be automatically doubled- one portion to victims, one portion to the public lawyers doing the actual defending. But then the funding might get cut just because they'll see these payouts as a substitute or worse push a perverse incentive for them to let these BS cases continue for larger payouts eventually.

more like texas gave samsung a billion dollars in incentives to operate there... if any state offered a BILLION in incentives to keep 1,000 employees... they deserve to burn down

when you don't collect taxes, curious where all that money will be coming from

...essentially did that when Microsoft had Chinese contract engineers IN China working on US government infrastructure a little while back. But don't worry, American supervisors were monitoring them.

- The B2 bomber plans was when a lead designer/engineer on the project defected with a drive full of data
- The F-22 and F-35 programs though were severely compromised and TB's of data exfiltrated.... ...don't worry though... the US and China have a mutual understanding to not do cyber espionage on each other ;) ;)

that OPM hack was a complete shit show... setting back human intel gathering operations by decades and compromising countless people... those OPM records had photos, biometric data, sensitive lie detector and security clearance info... including data that can be used to blackmail the people

In fairness to CISA - this was a contractor... and i doubt CISA encourages anyone sync CISA stuff to their personal GitHub.

I've worked at many places where leadership did everything right, had the correct policies and DLP, etc... yet still... some jackass will have creds in plaintext somewhere it shouldn't be... chats, photos, screen shots, coppy/paste buffer, notepad, stickypad, email, browsers, etc... Not gonna blame the entire org for some dipshit that thinks they're above the rules and thinks their shit don't stink.

At some point, we do need to hold people accountable that were actually responsible for the data leak, and those that were supposed to and failed to detect this stuff. Wouldn't be surprised if this guy broke multiple internal policies, and laws by doing this. To work with fed computer systems, there are security practices you're supposed to follow. Publicly posting keys and creds to a 3rd party and disabling protection to that data making it public is definitely a NO NO. I also doubt CISA is allowed to scan GitHub or most other services for possible leaked keys/creds as it's against nearly everyone's TOS to do such a scan.

they've been receiving money from the governments (at multiple levels), and tax breaks, plus preferential access rights to certain land.... this was all for "public good"... so sure... cut people off from electricity... but also... refund the funds you ever received from governments (with interest and adjusted for inflation)... they were so eager to get those funds so they can "service" those areas and operate there... have them pay to remove every inch and piece of infrastructure, and have them restore the areas to their natural state.
F'it... be petty... bar them from operating in your jurisdictions or don't license them (new or renewal). Send the inspectors and fine them for every infraction. Send the tax cronies to audit the shit out of them.

no, was next on the night stand... right after i finish reading the source code to angry birds, I will be sure to definitely "read the chromium source code" next... :/