Posts by b0llchit 1914 publicly visible posts • joined 29 Mar 2011
...Mechanical Turk workers who were put to the test and managed a score of 60.3 percent. [...] "a 10.4 percent gap in overall accuracy remains when compared to the human baseline, leaving plenty of room for model improvement."
I'd say, it leaves plenty of room for improving the human population! These MechTurk people are supposed to have a high school degree and, as it seems, are not performing very well. Maybe that is why they work for Amazon.
No wonder people are afraid of "AI",... they are underwhelming themselves and are easily impressed by an ML mechanical turk.
Or maybe it is just the bell-curve striking again. Half of the population is by definition below average.
Or, as a cynic might put it, an attempt to ram AI down the throats of Windows users.
I'm not sure you need to be a cynic to classify the tactics as "ramming AI down your throat".
Maybe they should add a pop-up window on every mouse click with the question: "You are trying to select something. Please consider Copilot to click for you and improve your productivity. Would you like to continue >>[yes]<< / [no] and install Copilot?"
A third-party contractor running a database without password protection...
That should amount to criminal negligence.
and then: ...have been set to "public" in error, since access needed to be open to multiple organizations, including the police and towing and storage companies.
That is (criminal) gross incompetence at work. Anybody not applying a segregated and security-in-depth design must be mandated to pay for any and all possible and potential losses from now to eternity for anybody potentially exposed.
And the C-suite should be publicly flogged.
It is amazing how fast things become obsolete these days.
The "technological progress" is matched by how often we need to replace gadgets. Sure, modern stuff can be better, but the total cost-benefit calculation over a device's lifetime doesn't look good. I'd argue that the costs are higher, once you take total cradle to grave resource usage into account.
...discomfort with finding themselves dependent on Big Tech.
Wow! That must hurt to admit that they are a collective bunch of idiots trusting the untrustworthy with your data. As if the signs were not on the wall from the start and they hadn't been told by the knowledgeable. But being good politicians, they have a plan to shift the blame. They will blame "the other guys" and again may shine in the light of utter bullshit eaten up by other idiots.
However, Stocks said that, though significant, cost overruns were not his "core concern."
Please validate Brooke's law once more, add more people and throw more money at the project.
In the meantime, Birmingham will gradually transform into Oracleham. I'd say they are halfway through, but that isn't a "core concern" either.
Many projects live on the "submit a fix please" principle because they do not have any resources and don't have paid-for maintenance. When someone logs (many) problems, then that is very nice. But not adding to solving the problems is a real challenge.
It is "easy" to run analysis on code. But fixing the problems uncovered is often much harder because, most of the time, you need to understand the broader logic of the code before you can fix a problem or risk introducing other flaws. The resources required to find flaws is lower than those required to fix any found flaws.
As with most open projects, please get involved. It is appreciated when you report bugs. It is even more appreciated when you submit a fix for the bug.
But, that is brilliant! We must create a new programming language that cannot create bad programs. The applications are endless and huge. Just imagine, just for a moment, no more programs that can go wrong or do wrong. What a perfect world that would be. Surely, AI can help create this programming language?
We feel we must ask: couldn't this have been done as a kernel personality instead?
And risk insanity when you hit the limit on multiple personalities with so many distributions available?
...the flaws found in curl would not have existed had it been written in a more memory-safe language...
Instead, the logic errors would probably translate in denial of service problems.
A memory error is, of course, serious and problematic, but the cause is due to a logic error in the code. Like, I expected 3 but got 4. Rewriting the code in another language may very well fence off the class of memory errors. However, the logic error now translates into an exceptional state that needs to be handled. And how do you handle an error you didn't account for? Well, you can be lucky and somehow terminate or you can end in an endless loop. And that is a clear opening to a denial of service. One can argue that DOS is not as bad as RCE, but when it overflows/deletes your data stores, then you would not be happy either.
A feature to control the inside of the box from the outside is a recipe for disaster. We've seen previous examples like activex and com/rpc. But every new generation of code must apparently pass through the phases
Apart from obvious, you can ask Oracle to take over. I hear that they really would like to bankrupt a country after successfully taking over Oracleham. Better to be on the right side of the red line protection racket before they come after your count[r]y.
...to establish international AI regulations...
followed by ...we are working on international guiding principles and a code of conduct...
So, which is it? Mandatory rules or a "please don't do that" paper?
My guess is some voluntary guidelines because mandatory rules are rather futile. The ML/GPT systems are out there and you cannot put that genie back in the bottle. Now telling people "you may not use this for X and Y" is like demanding that the world stops spewing propaganda. Not gonna happen.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
