AI Security is Already Here

Image created by Google Gemini

An article was published late in the day on 16 September. It was originally slated for a much higher-profile technology site - as the feature - but leadership of one of the companies in question threatened them if they ran it, so it was shuffled along to a point of obscurity.

I find it amusing that people think these strategies are viable in the long term.

Our technology works, and we're making it available. We've solved "nearly" all of the problems (I only qualify that in the event something else happens to bubble to the surface) and are already well down the path outlined in my recently re-published Microsoft-Linux essay.

They're trying to fight, so we're not far from winning - again.

The correct place to fix a problem is the place where the problem enters. Not the place where the problem manifests.

Voxiferi - Stop Bad Data at the Door

History Repeats

 Mark Twain once said: "History doesn't repeat itself, but it often rhymes."

In the early days of the Linux revolution no one took us seriously. We were dismissed as rabble-rousers and seen as dreamers and delusional idealists. I like to view us as believers. In a time when the world was becoming more disconnected a group of people who leaned towards being anti-social (most nerds and geeks are neurodivergent to one degree or another) found common ground and a common cause. We were family-like in every respect; we'd have intense disagreements one day and celebrate someone's birthday the next - everything was about improving Linux.

In the spring of 1999 there was a disturbance in the force. This is an article I wrote and published. It still survives in the Internet Wayback Machine.

Today it holds renewed significance as we stand on the cusp of being re-branded rabble-rousers, dreamers, delusional idealists, and perhaps a few other adjectives. If our estimation is even remotely close we're destined for uncharted territory without so much as a compass. Things are about to get interesting.

Then You Win 

May 18, 1999, 02 :35 UTC 

 In recent months, I've had the opportunity to sit back and watch a prophetic statement made many years ago unfold before me. Many of you have seen the statement -- at Linux Hardware Solutions, we have distributed it on T-shirts since last October. The quote is attributed to Mohandas Gandhi (or Mahatma Ghandi, depending upon the reference used)1: 

 First, they ignore you. 

 Then they laugh at you. 

 Then they fight you. 

 Then you win. 

I'd like to provide my perspective on these few lines and how they relate to the relationship between the corporation known as Microsoft and the Linux community. I use Linux here, not to diminish any other moniker that might apply -- be it GNU, OpenSource, Free Software,or *BSD -- but due to the fact that Linux is the target with the highest profile and has attracted the attention of Microsoft. 

First, they ignore you 

I don't think we need to dwell too long on this point. After all, how much can you say about things that didn't happen? For the first several years that Linux was mentioned, the Microsoft response never amounted to more than, "What's that?" So, although a measure of the ignorance may have been feigned, I doubt there was more than a ghost image on their radar scope. 

 Then they laugh at you

My first recollection of this goes back to sometime last summer when Steve Ballmer made a habit of dismissing Linux as nothing more than a toy being used to amuse computer enthusiasts: students as well as professionals. This phenomenon built momentum, with Ballmer being replaced as spokesperson by the one we all know and love, dear old Ed "The Mouth" Muth. Ed Muth's role has metamorphosed from stage two to stage three, which we'll take a closer look at in a moment. Along with the participation of Ed Muth, we've had the dubious honor of attracting direct references from none other than Bill Gates himself. Although less frequent, the outbursts by Mr. Gates have followed fairly closely the tone of Muth. Until recently, Microsoft appeared to be stuck between this stage and the next, apparently trying to figure out who they could fight. They started tossing out real FUD one minute, then they would change to a more patronizing tone, apparently fearful of providing too much credibility to the movement due to their protests. 

Then they fight you 

The recent Mindcraft fiasco is probably the most poignant example of this stage. I think it actually started with the Halloween documents, since it is apparent in these that Microsoft is at least trying to figure out how to fight, even if they didn't know who to fight at that point. Judging from the Linux Today archives, an off-the-cuff, totally unscientific survey appears to indicate that the FUD campaign started in earnest sometime around the first of March. The Mindcraft "study" was an apparent attempt to add some credibility to the FUD from Redmond. I'm being charitable when I make the observation that it didn't quite work out as planned. Where will the next blow strike? What will be the next target? When all is said and done, will the answers to either of these questions really matter? Whether they realize it or not, Microsoft has attacked a hydra. If they manage to successfully sever one head, at least two more will take its place. Microsoft is finally pitted against a competitor that they cannot buy, cannot bankrupt and cannot combat in the traditional sense. They see their competitors embracing Linux, they see large, high-profile customers embracing Linux, and they don't know what to do. 

Then you win 

They're fighting, so it's only a matter of time before we can claim victory. The recent invitation to have Linux pitted against NT in the ZD labs is the most direct affront so far. We don't just have Microsoft publishing the results of a study, we have them issuing a direct challenge. If they didn't view Linux as a legitimate threat, why would they bother? Unwittingly, they have provided a level of promotion that could not be purchased otherwise. There is no need to walk into their trap. We can play this game on our terms, when we're ready. If we keep our wits, there is no way we can lose. Will Linux mop up the floor with NT across the board? Probably not. But Linux doesn't have to. Through the simple act of issuing a challenge, Microsoft has already crowned Linux the victor. 

 1. Yes, his name is Mohandas (Karamchand) Gandhi. He has also been known as Mahatma Gandhi and (incorrectly) Mahatma Ghandi. There are several references that I have located that list the first and the last, and other references list all three. That's why I mention in the paragraph at the beginning that there are references to different spellings of his name. Email comments are welcome, but I don't need anymore reinforcement on the spelling of his name. I listed the confused spelling as a point of reference and used his proper name in the main body. Thanks to Ajay Shah and Michael Hall for additional insight. 

 Kit Cosper started using Linux in 1993 and was a member of the Red Hat Development Team from 1995-1996. In 1996 he founded Linux Hardware Solutions which has recently joined forces with VA Research. When he's not at work Kit enjoys SCUBA diving, fishing and spending time with his wife and three children.

Hold on to your hat

 "The question of whether a computer can think is no more interesting than the question of whether a submarine can swim." -Dijkstra

AI is not intelligence, it is data processing parading as human reasoning. Given that human reasoning is inherently flawed, ergo not worthy of emulation, trying to force a machine designed for precision to implement human intuition is nonsensical.

I've been warning for years that AI was going to collapse as it began to ingest it's own output. The day has come, albeit much sooner than I would have posited. Over at Voxiferi a new law of IT has been discovered. It's been tested and proven, patents are pending, and opening hands have been played.

Garbage In – Garbage Out has been proven much more detrimental than just producing nonsense.

Stay tuned, it's going to be a wild ride.

Not podcasting after all...

Legally Profiting from a Ponzi Scheme

Our group was recently recruited to contribute to a start-up technology organization in a rapidly developing market sector. According to their narrative, they were sitting on a war chest of $1B in venture funding, and in the process of building strategic partnerships to accelerate growth rather than face delays developing expertise in-house. In our realm they were seeking assistance with system architecture specification and testing to implement their novel, multi-layered, distributed approach to a high-intensity compute paradigm. Their proposal promised an elegant, optimized approach showing great potential for being a disruptive force in the industry. The concept was simple; don’t just toss more memory and processing power at the problem, build a system that offloads different types of jobs to specific processor architectures, each ideally suited to the task at hand. Building the communication structure to handle this task would be no mean feat. Our team has participated in the design and construction of multiple large cluster systems over the years, delivered to various “Three Letter Agencies” of the proverbial “We could tell you, but we’d have to kill you” variety and other systems used in massive international construction projects – so we have a catalog of experience in the design of unique systems.

The start-up deployed significant lobbying and provided strong encouragement to sign-on one of our group as an employee, but that didn’t align with our long-range vision, and would have needlessly complicated our internal accounting while creating taxation issues. Maintaining a simple consulting relationship between business entities was much cleaner, allowed for internal distribution on the basis of contribution, and avoided over-complicating taxation issues. Imagine the nightmare if one person were employed and received compensation that then needed to be redirected to another in consideration for their work. A simple consultancy agreement mitigated everything.

A Consultancy agreement was negotiated and executed, reflecting a considerable discount on our standard rates in consideration for other non-cash benefits, such as shared marketing and promotional opportunities. The agreement was based on a flat daily rate with a framework covering the commitment of resources to the project. We could plan our schedule, and they could manage their cash flow; a beneficial arrangement for both parties. Our work began upon execution of the agreement. In the first month we hit the limits as we got up to speed on the abstraction layers and developing conceptual outlines for workload segregation, routing, and aggregation. As the month progressed, they invited a member of our group to join their C-level team in an advisory capacity. This relationship was agreed, so long as any direct compensation remain directed through the existing agreement. Having an industry authority listed among their leadership would certainly benefit the organization and, by proxy, our group. A rising tide does lift all boats, after all.

Per the terms of the consulting agreement, we submitted an invoice at the end of the month for our work, with the balance due upon receipt and payable via wire transfer. Two days later we were still waiting, with no commitment for when we should anticipate payment. As the delay continued, we deepened our investigation and discovered some startling information.

• Employees were granted contracts from a non-existent corporate entity.
• No one had received a paycheck. (Remember the claim of a $1B investment raise.)
• There was no corporate structure, only an LLC – with $1B invested?
• Fraud charges against a principle settled with the government.
• We were directed to build out a sandbox for testing the distributed framework that was abstracted in our initial round of development, only to discover that the hardware required for construction was not even on order. The lead-time for the hardware exceeded the deadline for initial validation. (Perhaps an attempt to manufacture a claim of failure to perform and extract damages from us?)

Our spidey-senses started tingling, and the organization was instructed to scrub any reference to our team from their website and/or promotional material. We requested an update on invoice payment and were met with various iterations of “we’re working on it.” ($1B in the bank and they’re having to “work on” paying an invoice that is a small fraction of the daily interest that $1B should earn.) We informed the organization that we intended to terminate the consulting agreement under the cancellation provisions due to non-payment of the invoice.

Several conference calls followed, with assurances that we were merely victims of “growing pains” and things should flow much smoother in future as processes were developed and implemented. With these assurances we withdrew our intent to cancel the agreement, but we remained vigilant. Over the next few days we were met with more news of payment delays. We redoubled our investigation, which revealed even greater irregularities.

• There were over 100 LLCs associated with the organization. Each LLC represented a different physical location.
• Investments were actively solicited in units of $5000 or less.
• Those considering larger investments were encouraged to diversify across multiple locations. (i.e. invest $5000 ten times instead of $50,000 once.)
• Projections of ~25% returns on your investment.
• Multiple lawsuits and settlements pertaining to prior executive behavior.
• Unauthorized, and rather abusive, use of our platforms to promote the organization, resulting in measurable financial losses.
• Claims of strategic partnerships with LARGE network hardware manufacturers were pure vapor, as confirmed by first-person contacts at these manufacturers with direct knowledge of these types of relationships. 

With the discovery of these troubling patterns, we redoubled our collection efforts and turned up the pressure. We were shuffled from one person to the next, with the promise at every transfer that “this” person had the authority to resolve the problem and have payment issued. After a day of playing “hot potato” we ended up back on the CEO’s desk, because apparently not even the CFO could pay an invoice (even with a billion dollars in the bank…)

Then came the coup de grace.

The organization was conducting virtual meetings/conference calls with an operating systems company that we have deep ties to. Some of us have known each other for decades, and relationships go far beyond the professional. Spouses and families are known on a first-name basis, vacations and holidays are common topics of discussion. To say these are friendships would be an understatement.

In the moments leading to one of these teleconferences our friends were waiting on an open, recorded, line. The CEO of the organization and his spouse were also connected, and others were scheduled to attend. 

Our friends overheard the CEO making plans to avoid paying our invoice.

It was recorded.

When their teleconference completed our phones rang.

Our friends let us know what they overheard, and that the conversation was recorded.

With the smoking gun on the table, we finally had the leverage needed to bring everything to a conclusion. When we presented the evidence to the other parties suffice it to say they were a bit rattled. They offered us less than 10% of the balance due to go away.

We laughed.

An agreement was finally reached granting us legal damages for their breach of the consulting agreement. We were sent photos of computer screens confirming a wire transfer from personal accounts (not a business account with over $1B – shocking.)

By our best estimate, even our generous concessions placed a significant burden on the personal finances of the CEO, since the massive venture funding contribution was non-existent the entire time.

The funds were transmitted by wire transfer and confirmed safe in our account.

In the weeks since we've received confirmation that the nominal CEO of the organization was boasting to his remaining Board of Directors that he ripped us off for dimes on the dollar. A direct violation of the agreement, and exposing them to greater liability.

We're assessing available options now, and will likely seek additional compensation.

And that’s how we managed to profit from a Ponzi scheme, legally.