In related news, corporations look for ways to replace overpaid, underqualified CEOs with AI systems, which is expected to dramatically increase profitability.

I redesigned the databases that the CDC uses to track lab data for Covid and similar (which happens in near-real-time). Without getting deep into the specifics: before, it was a bunch of custom integrations per US state, feeding into multiple incompatible databases. Lots of manual work to see the big picture. Now, it's been moved to a single search database, with a common model for all data, regardless of source. I created the common design. Integration engineers had to tweak how each state's data is transformed to fit into the new one-size-fits-all standard.

Note that the system could work with non-US lab data, if politics allow it, to provide a global perspective.

So, the next time there's a reason to snapshot full-country data, it's easy. E.g., for the summarized data given to CNN and similar, e.g., to turn into hot-spot maps when an emergency hits. The data is already full integrated nationally, and can be queried in probably just seconds. Since it's generalized, a completely new disease should quickly be tracked without difficulty. It doesn't solve the problem that the people in charge often have to make "informed guesses" on what the actual solution is, but at least they have immediate data to work with.

AI is just another software tool.

Without "AI", but just implementing automation, I've created software over the years to eliminate endless roles, mainly when employees did things that were overly repetitive.

Examples: A team of 20 accountants (I automated validation of sales versus bank and credit card records). A team of sales pattern analysts (using the tool JustEnough to forecast product manufacturing and stocking needs nearly a year in advance, instead of iffy spreadsheets). Most of the team that manually did billing each day (leaving maybe just 5% of bills to be tweaked manually when there was a red flag on anything). Autogenerating software design diagrams (UML) and test requirements for big military projects. And the painfully tedious task of figuring out credit transfers at an online university.

The part of AI that hasn't been addressed well: proof of why the AI engine gave the result it did. More than 20 years ago, I helped write AI software to give a patient-specific dosage of medications in hospitals, with an aim of 30% reduction in how long they had to stay (yes, it makes that much of a difference!). It was quite effective, but hard to prove to a degree accepted by the FDA, due to the complexity of the logic. The product never got past that, and the web site is now dead.

Today, ChatGPT and similar make it far worse, since it's far too "voodoo" how it happens.

Fortunately, I saw that a group is working on a data trail showing the underlying decision logic in AI outputs. But, I suspect it will be many years before an audit trail of AI becomes common, especially since many companies base their entire profit margin on keeping how things work (or don't - ChatGPT lies a lot!) a secret.

I saw a YT video the other day that showed that, while a family who went to college and purchased a house when they were affordable is comfortable living on $75K/year, the cost for a younger family with today's absurd price is closer to $175K/year. That would require being in the top 6% in the US. Average is salary is about $69K, and $95K+ is top 25%. So, when it appears that nobody can buy a house, pay for healthcare, get their children through college, and save up enough for retirement, it's absolutely true (!!!).

A couple of weeks ago, I was trying out some Xylitol gum. Soon, I ended up with some obvious side effects. Mainly, I had really bad swollen feet and ankles, which I assumed were an allergic reaction, even though it seemed like exactly like the classic effects of type 2 diabetes. I stopped using the gum, and the symptoms soon vanished (no other change in behavior or diet). Now I realize that Xylitol really was giving me instant type 2 diabetes, it wasn't an allergy. Yikes! I'm throwing my remaining gum away instead of giving it away.

My prediction is that Netflix will cut back even more on their US-based on-site productions, now that they are being openly and blatantly rejected by the actors and directors and such of Hollywood. That funding will be redirected or eliminated in several ways:

• Even more of their money will be sent to other countries.
• They will imitate how series like Disney's "The Mandalorian" are created, with the technique known as "The Volume". In that technique, digital technology and projectors eliminate both studio sets and on-site sets entirely, emulating the whole thing in a computer. This lowers significantly what is paid across the entire production, cutting back dramatically on paid hours for the entire staff, and eliminating some jobs entirely, including building sets, and most post-production work.
• Emulating more of the cast. Note that part of what allowed Lord of the Rings to be filmed was emulating much of the cast, all the way back in the late 1990s (using the software tool "MASSIVE"). It was mainly used for any scenes that showed an army. That technology is orders of magnitude better today.
• The recent strike by the Screen Actors Guild was in large part about blocking technology (e.g., "GPT") from eliminating their jobs. While they think they won, the "writing is on the wall" that less money will go to writers, regardless of the SAG rules.

Companies are using AI and robotics to leave everybody unemployed. They will soon wonder why their bankrupt former customers don't give them any money. Eventually, the bosses themselves will likewise be replaced by AI, since overpriced management is far easier to replace than skilled labor.

Note that when aphantasia was discovered by Francis Galton in 1880, he found that MOST of the scientists that he knew had it. So, while we're only 2% of the population, we are extremely overrepresented in engineering, sciences, and some arts.

I'm a classic example. I'm an engineer ("senior software architect"), and, as is common for aphantasia, I depend on creating extremely detailed technical diagrams of my designs, since I can't "see it in my mind." That led to gaining my roles. Also, typical of aphantasia, I dig deep into topics. For example, I created what was probably the first semi-automated software for the diagnosis of psychiatric patients (for the US NIMH). Twenty years later, I not only remember the topic, but have become a domain expert on it in many areas, and recently ended up working remotely w/ a grad student in another language to translate source materials for her Masters degree in psychology into her language to speed things along.

I have difficulties pronouncing other languages well, since I can't recall what they sound like, and can't even recall how I sounded moments after speaking to correct myself. But I learn languages fast. I became fairly fluent in Spanish in just two months. While many people figure out translations in their heads before they speak, I never had that step. Translation has always been instantaneous for me, since there is no concept of "planning what to say in my head."

The caveat is that I've always tended to be too blunt, be it in English or another language. I say what I think, instead of overly-rehearsed statements that are politically correct but untrue.

It isn't uncommon for people who have aphantasia to have an exception to the condition. In my family, those of us who have aphantasia (my mother and most of my siblings) have the exception that we can hear songs in our head, and become musicians. In my case, I play more than a dozen musical instruments and am a composer. I have a sister who was a successful professional singer, actress, and dancer for many years, even though aphantasia makes her unable to see the scenes and choreography in her head. She now manages national teams of performers, and the fact that she needs to put choreographies on paper is in the long term probably an advantage.

A huge bonus typical of aphantasia is we are immune to PTSD. I nearly died in a severe ("99.9% unsurvivable") car wreck while I was in college, and have lasting effects. I remember nothing from the accident, even though I was conscious at the time, and I recall little of the months of surgery and recovery. The 98% who don't have aphantasia would have ended up a mess in that situation, but I continued largely like nothing had happened. The story of what happened to me is very much as if it happened to somebody else, not something personal.

It didn't crash, it proved its value as a kinetic weapon for lunar warfare ;-)

No, of course, you don't!

In 1998, Borland was in serious trouble. Their "solution" was to rename the company to "Inprise". The result? People assumed Borland had gone bankrupt. Two years later, it was renamed back to "Borland", but the damage was done. By 2015, all remaining product lines had been bought out (and a rebrand of the buyer happened as well, now it's "OpenText").

If the record labels want to block ChatGPT music that has obvious imitations of existing songs, they open themselves up to the same scrutiny. Imitations have always happened, and you would be shocked how blatant it often is.

He broke the law. But a big company suspending accounts without explanation is also very shady. It happens all the time, and innocent people can suddenly lose their business without recourse. Transparency is essential. There shouldn't have been a need to bypass the system to find out what's going on.

Scarlett Johansson

Nuff said ;-)

Is it just me, or is charging $6K for maybe $0.60 of peanut power (a 10,000x increase) a steep markup?

It's always the classic of "first to market" can be a disaster when it involves a truly new technology.

Examples of when it goes wrong for a very long time (but if you throw money at it long enough it can work):

• The F-35 fighter
• The Samsung folding-screen cellphone

Apple gets complaints about copying features of their competitors, but they end up far more profitable due to avoiding putting a feature on sale unless it's truly ready (with the notable exception of the bad laptop keyboards about five years ago).

The company that got away with selling a feature that isn't ready is Tesla, by Elon being very clever about it:

• The self-drive feature initially was alpha code at best, but people paid for the promise.
• Working features were added little by little. The system has far more capability than it offers to the driver, since it is running tests of the latest alpha/beta software in the background without those features showing up in what happens as you drive. Yes, you pay extra to drive around acting as a beta tester.