not randomly firing people.

And it was clear that a chatbot was the better move.

I haven't looked at the code in years, but the last time I did, storing the codes meant that the merchant was responsible for any fraudulent charges in the event of a data breach. I've mercifully not had to store credit card data for a few years now. Back then it wasn't strictly prohibited, just a really bad idea from a risk management perspective.

To me, this means that the CDC has reason to fine Delta Dental for negligent exposure of PHI.

If this is the case. Delta Dental could be subject to a fine of up to $482 billion. is a reasonable primer on how bad the fines could get.

It is simple at small startups: the person doing the firing coordinates with the person who is going to pull access at the beginning of the termination meeting the access is pulled and the person pulling access texts the person doing the firing when access is revoked and it is okay to end the meeting.

Now if you don't have a list of your critical accounts.... well you have bigger problems, but you can probably at least figure out that the password manager, AWS, and Google access needs pulled.

More issues is a more or less bogus statistic that Consumer Reports has.

If you have a car with 1,000 features and 10 of them have issues and another car has 12 features and 8 of them have issues consumer reports would consider the car with 2/3rds of the features failing as more reliable. Yes, the numbers I'm giving are exaggerated but they illustrate the point.

It made sense years ago when cars had a similar number of features, but now that the number of features varies by orders of magnitude it doesn't make much sense.

Consumer Reports has some strong biases and has skewed tests to arrive at predetermined outcomes. (The Suzuki Samari being one of them, they modified the test until they rolled the car even though it survived the first dozen or so attempts.

They could have just said that cars with more features have more things go wrong with them.

The question the researchers aren't answering is that they assume it is true data, but what percentage of the "training data" is fictional data made up by the model?

They have only proven that the model spits out things that look like it might possibly be training data, but could also just be probablistic strings that are not in the training data. I'm not sure how they are proving that the data exposed is real.

No clue.

But, what I do know. This is patent law and not copyright. You can't patent natural laws. You can't patent math, but things like RSA were patented so the line isn't super clear. I miss the old days of slashdot where we could have counted on a well known patent attorney chiming in.

I'll argue the other side. With every upgrade that fixes bugs, it is just a matter of time before those bugs are turned into CVEs that require upgrading.

As such, developers need to design their code to work with shifting versions of dependencies, and CI/CD systems need to be designed so that they can adequately test code so that they can reliably run updates in an automated manner. If you fail to do that, your software will eventually be viewed as unusable by internet connected computers.

Let me elaborate. Most office buildings are purchased with interest only loans that have balloon payments after five to 10 years. The financing is based on the rents that the building generates covering the interest payments.

Anyone paying attention to interest rates can see where this is going wrong. Interest rates are roughly twice what they were five years ago, so to cover the increased interest rates rents would have to double to refinance to make the balloon payment and rents are down, so people are walking away and losing their down payments.. I'm not sure I understand the logic of giving those loans at 3% in previous years. Probably theory would have said you had at least a 20% and maybe 30% chance of losing money for a 1% increase in your yield

Or that people who buy $2,000 mattress toppers in San Francisco are a specific demographic that has a large intersection with the people who work in AI. As there are thousands of people in San Francisco who work in AI in San Francisco

How about an Apples to Apples comparison:

• BMW M3 weight: 3,840 to 3,990 lbs
• Tesla Model 3 weight: 3,862 to 4,048 lbs

That puts the EV less than 100 lbs more than an ICE car which is roughly the same. (performance and size wise)

I'm not sure that the "EVs are heavier" stands up to closer scrutiny. Modern cars with all of their safety and convenience features are heavy

They could spend all their time trying to get GCC to constantly compile decent code on Itaniums. Either that or create a new language and compiler that makes it possible to write performant code on Itaniums.

Snow slowly sublimates in the dark even when below freezing.

This is different.