Ray argues that AI does not hallucinate, but instead confabulates. He explains the difference between the two terms:
“A hallucination is a conscious sensory perception that is at variance with the stimuli in the environment. A confabulation, on the other hand, is the making of assertions that are at variance with the facts, such as “the president of France is Francois Mitterrand,” which is currently not the case.
“The former implies conscious perception, the latter may involve consciousness in humans, but it can also encompass utterances that don’t involve consciousness and are merely inaccurate statements.”
And if we treat bots (such as my Bredebot) as sentient entities, we can get into all sorts of trouble. There are documented cases in which people have died because their bot—their little buddy—told them something that they believed was true.
Adapted by Google Gemini from the image here. CBS Television Distribution. Fair use.
After all, “he” or “she” said it. “It” didn’t say it.
Today, we often treat real people as things. The hundreds of thousands of people who were let go by the tech companies this year are mere “cost-sucking resources.” Meanwhile, the AI bots who are sometimes called upon to replace these “resources” are treated as “valuable partners.”
Are we endangering ourselves by treating non-person entities as human?
“ROC is an independent American artificial intelligence company redefining the global standard for Vision AI in identity, security, and digital forensics. Our Vision AI platform delivers real-time facial recognition, multimodal biometric verification, video analytics, and AI-powered evidence analysis to mission-critical organizations across both private and public sectors. ROC’s biometric algorithms are routinely ranked by National Institute of Standards and Technology (“NIST”) as among the most accurate and computationally efficient globally. Our solutions outperform legacy foreign-built systems at a fraction of the cost, with faster deployment and stronger trust. As demand for trusted AI accelerates across law enforcement, defense, and regulated commercial sectors, ROC is scaling rapidly through a growing network of integrators and multi-year deals. We are expanding from a foundation of government leadership into high-growth commercial markets such as access control, physical security, and identity verification. Our international pipeline spans the Middle East, Asia–Pacific (“APAC”), and other strategic regions where national AI and identity investments are surging. With sovereign U.S. development, deep technical leadership, a vertically integrated platform, and proven field results, we believe ROC is positioned to become the category-defining leader in operational Vision AI.”
Of course, as ROC enters new markets, its “made in the USA” strength could potentially become a weakness. For example, EU regulators may (or may not) become wary of using algorithms from a non-EU company. And forget about Russia.
Any SEC-governed statement must detail risks to ensure that investors are not misled, and ROC lists the types of risks that one might expect (dependence upon certain types of customers, complex product lines, etc.). But this particular risk caught my eye:
“If we are unable to successfully deploy our marketing and sales organization in a timely manner, or at all, or to successfully hire, retain, train, and motivate our sales personnel, our growth could be adversely impacted.”
Hey Scott…
Also buried in the preliminary prospectus are sales and cost figures from 2023, 2024, and the first 9 months of 2025; a note that two customers accounted for 45% of ROC’s revenue in the first nine months of 2025; the negative consequences of ROC’s mission “to support Western liberal democracy and its strategic allies” (companies with fewer scruples can sell to all sorts of entities that ROC won’t touch); risks related to artificial intelligence (not an issue when Printrak went public in the 1990s); and many, many, many other risks.
Because that’s what you say when you want to go public. You always, always use caution when talking to investors.
So we’ll see what happens. Is this a good time for an IPO?
The Institute for Defense and Government Advancement (IDGA) recently released some survey results. Now I don’t want to simply reproduce the results; go here to download your own copy of the report.
But I do want to say this.
“A large number” of IDGA survey respondents expressed concern about “Interagency information sharing.”
This is NOT a technology concern. The technologies exist to enable information sharing. For example, one of Bredemarket’s clients recently made the technological changes necessary to allow an application, designed to interface to agency A, to instead interface to agency B.
No, this is a business concern—or in this case a governmental concern. A matter of setting up the processes to allow Bob from agency A to exchange data with Judy from agency B. Even though Bob thinks that Judy is a bozo, and vice versa.
And while we’re on the topic…
If you’re worried about Big Government (the FBI and the CIA and the BBC, BB King, and Doris Day) (or INTERPOL and Deutsche Bank, FBI and Scotland Yard) combining all their information to entrap you, your fears may be difficult to realize. Yes, there are cases in which the agencies share data. But there are also cases where they don’t, because it’s in an agency’s interest to keep its data to itself.
Agencies usually ask the question “How can I GET the data from the Bureau of Stuff?” They normally don’t ask the question “How can I GIVE my data to the Bureau of Stuff?”
And that’s why agencies run into problems sharing data.
“Perhaps the most visible change is the push for phishing-resistant authentication—methods like passkeys, hardware-backed authenticators, and device binding….This shift signals that yesterday’s non-phishing-resistant MFA (SMS codes, security questions, and email OTPs) is no longer enough because they are easily compromised through man-in-the-middle or social engineering attacks like SIM swapping.”
Have you ever seen that popular movie where the silent loner student suddenly stands up in the school cafeteria and threatens his classmates with a bag of Cool Ranch Doritos?
“After football practice Monday night, Taki Allen chatted with friends outside Kenwood High School while munching on Cool Ranch Doritos. When he finished his snack he put the bag in his pocket. Minutes later, several police officers pulled up, pointed their guns at him and yelled for him to get on the ground, he said.”
So why did Taki (I’ll get to his name later) receive police attention?
“The false alarm was triggered by Baltimore County Public Schools’ AI-powered gun-detection system, Omnilert.”
Yes, it…um…appears that the AI-powered system thought the Doritos bag was a gun.
“In this case, Omnilert’s monitoring team reviewed an image of “what appeared to be a firearm” on the person at Kenwood Monday night, said Blake Mitchell, a spokesperson for Omnilert.
“”Because the image closely resembled a gun being held, it was verified and forwarded to the Baltimore County Public Schools (BCPS) safety team within seconds for their assessment and decision-making,” he wrote in an email.”
Although not explicitly stated, it appears that the image was sent for human review…and the human thought it was a gun also.
So how can a Cool Ranch Doritos bag look like a gun? Let’s see the picture.
“Mitchell [noted] that their privacy policy prevents them from sharing the image.”
I probably spend at least 30 minutes a day playing smartphone games, which means that I probably spend at least 15 minutes a day watching ads.
I’m convinced that “games” are in reality ad delivery vehicles whose sole purpose is to lure you to download other ad delivery vehicles.
But back to the topic. I’ve seen some ads countless times, which means I’ve seen this misspelling countless times over the last few months.
Source not revealed.
And apparently no one at the company (a U.S. based firm) has noticed yet.
They would…if they reviewed their existing content on a set schedule with regular content review cycles.
This is a trick I picked up in the proposal world, one I try to implement when possible.
But not always. Apparently some people are still downloading my old SIX content questions from somewhere on the Bredemarket website. Gotta track that down and fix it.
Bredemarket 