TAM - part the eighth

Ah well... the noose gets tighter. Couldn't happen to nastier set of guys. Not many gals involved? New reporting supporting the disappearing TAM, if there ever were any, from AI, as currently implemented. How to piss away a Trillion-plus Bongo Bucks. And these are the guys who've convinced themselves, and a hoard of Daddy Warbucks clones that it's true. It ain't. The crash, one hopes, won't be as widespread in the economy as The Great Recession. One hopes.

In survey after survey, executives and their employees report unimpressive gains from A.I. tools. Researchers from the Federal Reserve Bank of Atlanta, Stanford University, the Bank of England and the Autonomous Technological Institute of Mexico recently published a poll of nearly 6,000 senior business executives in which about 90 percent said the technology had no impact on either employment or productivity. That's even though 69 percent reported that their firms actively used A.I.

Regular reader may recall the proposition that AI, as currently implemented, will benefit from RDBMS storage of ground facts, either a priori, or as adjudicated output from AI, as currently implemented. One suggestion was to simply store weights in a SQL (yes, I know, it's only meh relational) dztabase and use them to pre- or post-validate an output. We'll see. May be.

But A.I. is also creating a new type of work that may be sucking up much of the time the technology itself is saving.

Call it a "verification tax," "botsitting" or "rework" — it's the often arduous task of checking that A.I. output is accurate, safe and usable, and it is putting a very human bottleneck on A.I. automation.

This ohe is so, so precious

Tech companies, for example, can use A.I. to generate thousands of lines of code an hour. Yet Amazon and others still require a human to review that code.

IOW, AI, as currently implemented, is little more than a fresh graduate from The Grace L. Ferguson Airline (And Storm Door Co.)^©Newhart COBOL programmer school subsidiary. I guess AI, as currently implemented, coding does go a bit faster. Ya still need to keep a vewy, vewy close eye on the little devil.

To be a bit more generous: back in the olden days (in which I existed shortly before they went poof) of computer systems, the world was divided into two parts, coders and systems analysts. The former banged out the code, based on structure and instructions from the systems analysts. Said bevy of analysts had the final say on, and responsibilty for, the buggyness (hopefully, nil) of the delivered product. Now, with AI, as currently implemented, everybody is an SA; in those olden days, this was the position to which one aspired (unless one were a crafty techy who envisioned himself as Management Material). But, of course, back in those olden days, the heavy lifting (in terms of time requirements) was done by the coders and output rarely ran as high as 100 lines of COBOL a day per coder. Piece of cake. The SAs sat around bullshitting and bitching about how dumb the coders were. "Can't they just follow directions, fur cryin out loud????" And such.

And, in keeping with the mantra: since AI, as currently implemented, is built on all the crap coding festering in the innterTubes, building smart datastores and simple code to leverage them isn't likely. Remember the coders' mantra: "all the data be mine, and only mine, and only my code can make sense of it." That shit is what permeates the innterTubes. `

Old Wine, New Bottle

Well... Margaret Hamilton died. No, not the witch who pursued Dorothy. This is one of many who made the moon landing, and later ascent, possible.

She joined the MIT Instrumentation Laboratory, which developed the Apollo Guidance Computer for the Apollo lunar exploration program. Hamilton was initially hired as a programmer for this process but moved on into systems design.

What's notable about all this is just - up your butt Elon - it was done 60 bloody years ago!!! In case you've been living under a rock for a few years, not a bad idea now that Bone Spur Samurai^© is threatening all out war in order to get more climate change fucking up oil in Iran (and the rest of the Middle East, eventually), then note what Ms. Hamilton did

Her skills led to her promotion to director of software engineering on the project, which developed the two 70-pound onboard computers for the Apollo flights: one in the command module, which carried the astronauts from Earth, and the second in the lunar module, which set down on the moon's surface.

IOW - the hardware and software that allowed the lunar module to land ass first on the surface and fly off again. 60 years, more or less, before Elon and his acolytes boast about doing the same thing with LEO engines. And she didn't have all the compute that Elon has.

Developing its software was a major lift in an era when computers had limited memory. The onboard computers contained only 72 kilobytes of storage memory (an iPhone 16, by comparison, can have more than seven million times as much storage).

There are, appropriately, lots o reports out there. Of the many I read, only the NYT story mentions the landing coup. The others stress a fail-safe block of code that came in handy on Apollo 11.

This, if you're interested, is the computer.

The AGC was designed at the MIT Instrumentation Laboratory under Charles Stark Draper, with hardware design led by Eldon C. Hall. Early architectural work came from J. H. Laning Jr., Albert Hopkins, Richard Battin, Ramon Alonso, and Hugh Blair-Smith. The flight hardware was fabricated by Raytheon, whose Herb Thaler was also on the architectural team.

Mission Control was in Texas, but the real brains were in Cambridge. Elon you're just a poser.

Also, of some interest, the NYT report says that they built the software with 1s and 0s, implying that they worked in machine language on the metal. That would be so 1940s. According to the wiki page, of course, there was an assembler for the AGC. As one would expect by that point in time - all real computers had an OS and, at least, an assembler (may be not a macro assembler).

TAM - part the seventh

Yet another report from the NYT, which you might be able to see.

But basically, the idea was something like: These systems are merely good at seeming clever and are not going to be even capable enough to do a ton of economically valuable work.

AKA, TAM ≡ 0. Ha Ha Ha

This is the original from which that quote was built. It's not shown in quotes, and the original is behind a paywall, so it should be a precis of the original. But being behind a paywall... OTOH, the wiki has a page devoted to stochastic parrots. The term means what you think it does.

Again, from the NYT report

I thought: This is going to be a useful tool, and it's going to be powerful for a lot of things, partly because I believed then it would continue to get better. But I didn't think that people who were worried about more catastrophic scenarios were right.

He's changed his mind.

Two conflicting notions: 1) AI/LLM/whatever will never produce anywhere near enough end-user revenue to offset the trillions of Bongo Bucks already spent and 2) Ayatollah 17% Don^©'s newly annoited AI czar has surely been given the remit to take AI for Batshit J. Fucking Moron^©Tillerson's control of the United States of Alabama. Just as I, and ever more others, have been warning for a while. If you disagree: why wouldn't he? He does take whatever he wants, under threat of annihilation. This is the crown jewel of Dictatorship. He's going to pass it up?

Clayton unqualified for DNI, and unqualified for AI. Well... appointing a Gumby is kinda what he do. He's just a Gumby his own self.

I Told You So - 2 October 2026

Well... may be. Turns out, the AI hacking community does know SQL

In another case, the agents used what is called a S.Q.L. injection attack — where unexpected information is injected into a database in an effort to break through the network's defenses — to burrow into the systems of a company that operated a call center. They could then listen in on any of its customer calls.

Naturally, there's no guarantee that these AI engines either know how to build 5NF databases or care to. Since they're, so far as anyone knows, driven by kiddlie coders, they're most likely trodding the same RBAR world as their COBOL slinging GrandPa. Oh well. One might speculate that the "open weight" models have sense enough to store such weights in a RDBMS. Time will tell.

TAM - part the sixth

This is starting to feel like shooting ducks in a barrel. (I wonder whether the kiddies even know that sentence???) Yet more reporting on AI, but from a distinct perspective: ok vendors, you've gulped from the AI punch bowl to make a cheaper (may be?) and/or better product. God damn it, give me a lower price!!! I don't give a shit about your market share, only my wallet's share.

Now that AI is firmly entrenched in the business of killing off every last God damned white collar worker on the planet, let's have us some Shakespeare.

The first thing we do, let's kill all the lawyers.

Well... apparently not first, but they were the first to suck Ayatollah 49% Don^©'s dick, as you may recall.

First, of course, the ambulance chasers boast about embracing AI

Outsiders often cite law as an industry under threat from artificial intelligence. But the nation's biggest firms are eager to show off how they've embraced the technology.

After OpenAI announced a version of its latest A.I. model for lawyers last week, for example, a handful of top-tier firms proudly unveiled new tools.

Then, they get hoisted on their own petard; karma's a bitch

Some clients are using A.I. to make their argument for A.I. markdowns. Jennifer Leonard, the founder of Creative Lawyers, which advises law firms and corporate legal departments on adapting to A.I., said some law firms and their clients used A.I. agents to spot line items in bills where A.I. could have saved hours. "They literally have struck that out, and say, 'I want a discount on this, we don't want to pay for this part of the work anymore,'" she said. "Or, 'We know you have this new A.I. tool in house because we saw your splashy press release.'"

How will they pay for that summer house in the Hamptons???

I Told You So - 22 September 2026

Well... it's been rather a long while since I took up the mantra: "It's the Senate, stupid!!" The logic to that being, naturally, that gerrymanders don't (at least directly) affect state wide votes.

Now, the great and powerful Nate Cohn more or less agrees.

James Talarico has led every recent survey of the Texas Senate race, with polls showing him ahead among Hispanic voters by at least 20 points. Although his lead is often mostly dismissed as a reflection of the profound flaws of his Republican opponent, Ken Paxton, the polls also show the Democrat Gina Hinojosa to be surprisingly competitive in the race for governor against the incumbent, Greg Abbott, who leads by an average of three points in recent polls.

Now, of course, TX is, as always, an anomaly but nevertheless points to the obvious: state wide is easier to defeat than the gerrymander. There also remains the chance that the gerrymander in TX will come up short, so to speak

In a sign that the G.O.P. itself doesn't consider the races safe, Republican groups are spending in all six of these seats — which include four of the five districts Republicans thought they could flip with their new gerrymander.

Chocolate sauce on your pudding.

The Democratic Party's chances of having a truly exceptional midterm election — one that yields a large congressional majority and control of the Senate — could depend on whether Democratic gains among Hispanic voters translate into breakthroughs in Texas and Florida.

TAM - part the fifth

Well... the news keep on coming. It is my duty to carry on, when the news fits my agenda, of course. This new report on the angst amongst the Moneyed Class doesn't explicitly cite TAM for the tightening of sphincters amongst that class, but one need not read too deeply between the lines to see what's driving the angst.

But some investors are voicing increasing skepticism about the growth expectations for data centers and the risks associated with their build-out, forcing industry executives and their advisers to recalibrate their plans to raise tens of billions of dollars in public markets.

That bit, "growth expectations", is just another way to bleat "what's the TAM going to be when this shit goes live???!!!???" Really now. Who's going to pay out of pocket for AI porn when there's a torrent of free stuff? Not that I've dined at that table, mind you. But investing, and even staying in business past the expire date, depends on customers buying your shit. If anyone has a detailed justification for a TAM worthy of multiple trillions of Bongo Bucks investments, I'm all ears. Like Clark Gable.