mamund tumbles

Years ago, as a young jazz musician, i was told "You can only play the way you practice."

I later learned this was a variation on a quote attributed to Japanese sword master Miyamoto Musashi: "You can only fight the way you practice"

In the programming world ...

Pragmatic Publisher's Dave Thomas knows all about this, too. he wrote about it twenty years ago:

and Chad Fowler (a fellow saxophonist) responded in kind:

Unworthy kings would be tormented by the ancestors with false prophecies and faulty advice.

SILICON VALLEY, USA (Present Day) Yuri arrived at the office just before dawn. He liked doing this. Latte in hand, he would have a couple hours of "alone time" to work on the algorithm that drove the Semantic Assertion and Response Application (SARA). Few people understood the complexity of preparing and running these demos and today's was important. Investors from around the globe, as well as media, would be assembled in the main auditorium. They'd have lots of questions to ask and it was Yuri's job to see that they were answered satisfactorily. A disastrous demo could undo all the careful marketing and positioning the product team at Hjarna Systems had created. SHANG DYNASTY, CHINA (1600-1050 BCE) The daily divination was at hand. Oxen had been slaughtered. Shoulder bones extracted, stripped clean, washed, and drilled with tiny pit-holes. The prophetic pyre stacked high, lit, and properly fed. Hot irons were thrust into the heart of the fire in final preparation. All this before the shamans appeared at dawn to make their queries to the ancestral spirits. Yuri was used to this kind of pressure, he and Hjarna's CEO, Magda, had done this dozens of times. It was a parlor trick, really. Someone in the audience would ask a question, Yuri would re-compose the question in a form the bot would understand, enter it, and the results would appear on the display screen at the front of the hall. When all was ready, the King was alerted, carried to the great hall, and placed on the high seat to await the diviners procession to the central fire pit. Questions were posed. "Will it rain in this moon cycle?" "If we send 3000 into battle, will we be victorious?" "Will the King have a son?" "What's the future of AI?" "If we invest three billion in this company, will be achieve market dominance?" "Will Hjarna Systems have a successful IPO this year?" As each question was asked, scribes selected one of the prepared ox shoulders, the question was carved on one side and the bone was handed to the priests who made the bone "speak" by inserting hot irons into the pit-holes until the scapula cracked. It was this crack that held the answer. The shaman, sometimes the King himself, would "read" the cracks and pronounce the answer.

"No rain today." "The battle is yours." "There may be a son in the future."

What the audience didn't know was that Yuri had a staff of two or three assistants who also had instances of SARA running on their laptops. The group conspired to all submit slightly different versions of the question to the engine. Whichever one finished first (and had a positive reply) was the one that was displayed in the hall. That way, even if SARA crashed on a question, it was unlikely anyone would notice. The priests understood that a successful oracle knew how to phrase the question. The ancestors were never wrong, but the question might be. Ancestors were also known to mislead or confuse undeserving priests or kings as a revenge from beyond the grave.

Another trick the Hjarna team relied upon was volume. Initially, SARA's responses were short and direct. "The future of AI is murky at best." "Market dominance by a single company is unlikely." "IPOs for companies as young as Hjarna are rarely successful."

No one likes these kinds of replies. Instead SARA was engineered to spit out about five or six hundred words in each reply. To play both sides of the story, and to try to end on an upbeat note every time. This usually worked just fine.

Unworthy kings would be tormented by the ancestors with false prophecies and faulty advice. Their reign shortened as a result.

Here's another episode in the ongoing series "Ask GPT Chat" ME: Explain this quote: "Now I am become Death, the destroyer of worlds" in the context of generative AI GPT Chat: The quote "Now I am become Death, the destroyer of worlds" is a famous line from the Bhagavad Gita, an ancient Hindu scripture. It was famously used by physicist J. Robert Oppenheimer, who led the team that developed the first atomic bomb, to describe his feelings after the successful test of the bomb.

In the context of generative AI, the quote can be seen as a warning about the potential power and consequences of artificial intelligence. Just like the atomic bomb, generative AI has the potential to drastically alter the world we live in, for better or for worse. AI can be used to create amazing things, such as new forms of art or even help solve some of humanity's greatest challenges. However, if it falls into the wrong hands or is used without proper safeguards, it can also be used for destructive purposes.

Thus, the quote serves as a reminder that with great power comes great responsibility. We must be careful about how we use generative AI and ensure that it is used for the greater good of humanity.

she sits atop the rubble clutching a bloodied hand protruding from the smashed concrete and splintered wood her cheery playdress covered in gray soot her face vacant, voice silent

she is lebanese she is israeli she is palestinian

she is iraqi she is kurd she is pakistani she is chechen she is afghani

she is malaysian she is filipino she is cambodian she is uyghur she is somali she is congolese she is algerian she is colombian she is peruvian

she is ukrainian

"We're shutting down your project," she said, tossing my file on her desk.

"But I need just a few more weeks."

"No, after this last stunt, you don't need a few weeks. You need to resign from the collective."

I exhaled and stared at the floor.

"But I'm not recommending that. Yet." It sounded like a threat.

"My latest tests are encouraging," I offered. "It will work this time."

"Encouraging? Ha!" She grabbed my file again and rifled through the pages.

"Look at the history of your repeated tests. In your own data, the simulation results predict global nuclear war 3 times, genocide 7, climate disaster 4, community collapse 5, and not one round results in interplanetary colonization."

"But I've corrected those problems now. It'll work now."

Administrator rolled her eyes.

"Look, Rels", she lowered her voice, blew a few bubbles, and looked me in the eye. "We all admire your work on modifying growth codes to improve evolutionary speed and diversity. But clearly you've hit a dead end."

"We're so close!" I responded.

"No, you're not. Look, we all agree that your idea of tuning growth codes to increase aggression and competition clearly speeds evolution and learning…"

"Yes, that's my point…", I interrupted.

"But this overt aggression means the long-term viability of the community is greatly reduced."

"You've gotta face it," she said. "It's your choice of the genetic host that dooms your work. In all our space travels we've never found an example of an advanced life-form that came from primates." There was a slight sneer in her voice when she said the word "primates."

"So that's it?" I sighed.

"Yes. Turn in all your research and shut down your lab by morning. And, don't forget, I want all your growth codes from the similator, too." She'd calmed down now.

"Yes, administrator." I swam toward the portal.

Little did she know I'd already loaded the GC samples onto my private shuttle in preparation for a launch this evening. I just had to find a suitable planet w/ an existing primate population. One where I could easily introduce these modified GCs directly.

He lied.

A lot.

So often and so well that he could no longer discern fact from fabrication.

He was, therefore, endowed with a comfortable advantage when debating his opponent.

Unfortunately, he was also saddled with a predilection to grin incessantly.

Voters don't like it when you smile too much. It reminds them that you're lying.

Which he was.

But he'd forgotten that.

So he was confused when surveys indicated the public didn't trust him.

"Don't worry," his campaign manager said. "Your opponent smiles, too."

I was still pretty disoriented; a common side effect of 100 years of hyper-sleep, they told me. Trying to get a handle on my surroundings, I started asking my guide some questions.

"So, do you still have Facebook?"

"Facebook?" She seemed puzzled at the question.

Surely, in this bright and shiny 22nd century they must have something like #Facebook or #Instagram or #Twitter or even #WhatsApp, I figured.

"You know online social media. A way to keep in touch with friends, catch up on trending topics. That sort of thing."

She smiled in recognition. "Oh, social medication. No, SocMed is controlled like any other addictive substance. Only a few psychiatrists use it now. And rarely." She wrinkled her nose when she said the word "psychiatrists."

"Really?"

"Well, Defense Cadre uses SocMed for interrogations."

"What?"

"Sure. After 30 minutes of scrolling through a properly curated news feed, people are willing to belive any accusation or conspiracy you concoct and ready to turn on their own family."

"Huh." Now it was my turn to experience a moment of recognition. I'd never thought about the idea of Facebook as a torture device.

blood is thicker than water

and

After thousands of miles of alternating abuse and neglect, Wheelie made a break for it at ICT. To their surprise, no one seemed to notice them missing until some days later when they failed to appear at SMF as expected. By then Wheelie was already relaxing outside of STT.

blood is thicker than water

and

R: I'm not paying this bill.

D: That's not a bill, its your credit card statement.

R: Don't care. Can't afford it.

D: You plan to declare bankruptcy?

R: No, just not going to pay.

D: Why did you buy all that stuff last year?

yes, the latest use of the GPT engine has been applied to text content. ingest tons of words, categorize, templatize, and spew.

what we have now is a much more verbose, complicated "Magic Eight Ball".

VIRTUAL MACHINE

while the text responses are insipid, they point to some interesting possibilities. recently someone claims to have implemented a "virtual machine" -- now that's interesting! if this turns out to be true (and repeatable), there are some useful applications to this kind of work.

consider that gptVMs are really domain-specific worlds. they "know" about their subject matter and can offer up responses based on the rules within that domain. a linux VM is one example.

FASB ACCOUNTING

how about creating a gptFASB world. one that can "speak" and "think" in FASB accounting. once the GPT understands the basics, we _should_ have a handy bookkeeper at our beck and call. we just need to input transasctions (debits and credits) and ask for output as in....

"FASB, please produce this month's Balance Sheet."

"FASB, show me the Accounts Payable Report."

"FASB, please produce a P&L for this month."

and so on...

LEDGERS

of course, gptFASB operates on data kept in a *ledger* (ha, did i get your interest there?).

if gptFASB can operate on a ledger of debit and credit messages, how far off can a GPT that can handle Kafka EDA messaging implementations?

and if gptEDA is successful, it surely must have figured out MQTT, and/or HTTP, and more. now we can see engines that allow us to say things like:

PROTOCOLS

"Send this message via HTTP to the server."

"Filter the list of incoming data and route all sales topic messages to the sales-instance using MQTT." and so on.

i suspect this offers a rather new (and possibly lucrative) way to implement LO/NO-code solutions.

GPT can graduate from carnival attraction status to helpful tool with the proper corpus.

WIRES AND LIGHTS