TON Strategy Company received another Nasdaq “reprimand.” The firm moves fast and ignores some Nasdaq rules. Manny Stotz, the executive chairman of TON Strategy, has a number of balls to juggle. Nasdaq rules seems not to be one of the balls. You can learn what’s happened and why delisting may be a concern. The write up uses Bruce Lee’s arrow trick as a metaphor for Mr. Stotz’s actions. Why is Mr. Stotz channeling Bruce Lee. You can get some additional information in “Manny Channels Bruce Lee: Amazing Reflexes.”

Stephen E Arnold, June 25, 2026

Another dinobaby post. No AI unless it is an image. This dinobaby is not Grandma Moses, just Grandpa Arnold.

Absolutely the class action referenced in “Microsoft Shareholders Sue Over Allegedly Overhyped AI Performance and Hidden Cloud Slump” may be without merit. “Merit” is a slippery concept in today’s go go AI octagon. The issue seems to be related to marketing; that is, adding some sizzle to make that dried beef shingle worthy.

Thanks, MidJourney. Your art output reminded me of my exciting commutes on the 101. Rural Kentucky may not do the AI thing, but the traffic is definitely better than some AI systems’ outputs.

The write up says:

Microsoft is facing a class action lawsuit from shareholders accusing the company of defrauding investors by overhyping its AI products and hiding weaknesses in its cloud business. The complaint was filed in federal court in Seattle by the City of St. Clair Shores Police and Fire Retirement System, based in Michigan. Microsoft has dismissed the allegations as without merit and said it will defend itself in court. “Microsoft stands by the integrity of its public statements and will vigorously defend itself,” the company stated.

Okay, but a police and fire retirement system is likely to generate some media coverage in outlets from Police1 to the Telegram law enforcement channels. With Microsoft hoping to retain its grip on its US enterprise business, the Google is likely to perk up and monitor this legal matter as well.

The write up continues:

The shareholders allege that Microsoft portrayed its partnership with OpenAI as strong and stable despite ongoing fragmentation between the two companies. They also claim that Microsoft promoted Copilot adoption more aggressively than the results warranted. Additionally, they argue the company failed to adequately disclose or played down the financial commitments needed to build the data centers required for advanced AI models. The complaint further states that Microsoft did not properly disclose a decline in cloud revenue while continuing to invest billions in AI infrastructure without seeing proportionate returns.

What happens if one of the retirement system’s legal eagles stumbles upon the remarkable Ed Zitron? That individual has clawed his way (yeah, that’s a pun) to the top of the  “AI is a problem” heap of pundits. What if that same group of legal eagles get in touch with Dr. Gary Marcus. Those two could add some spice to the allegations that big, management-starved Microsoft is arguably better at:

1. Blowing its PR lead in AI to the estimable Google
2. Floundering around in the kiddie pool with OpenAI and a number of enlistees as other companies deployed more useful and reliable smart software to help people with routine office tasks
3. Pulling off one of those fancy moves that are the trademark of the tango champions Yanina Quiñones and Neri Piliu with the breath-taking “yes, it is Copilot and no, it is different” AI flourishes.

The write up notes:

The lawsuit is a formal legal challenge to Microsoft’s AI claims, which are among the more heavily promoted in the industry. If the case moves forward, discovery could reveal internal communications about Copilot adoption metrics, tensions in the OpenAI partnership, and infrastructure cost forecasts…. The case could become a reference point for similar lawsuits if other AI companies report disappointing results that fall short of investor expectations.

I don’t have a dog in this fight. Personally I am skeptical of the current crop of smart software systems. I do pay attention to quite specific applications of AI in certain policeware and intelware services. These are usually constrained and provide quite useful benefits to investigators working certain types of cases. But AI to design a treatment for my grand daughter’s illness? Sorry, none of the AI outfits have my confidence. I spent too much time analyzing the IBM Watson – Houston cancer thing.

Many interesting AI “actions” are underway. However, the combination of that living management case study in the Seattle area and a fire and law enforcement retirement system is different.

Net net: Hire Ed and Gary as expert witnesses.

Stephen E Arnold, June 25, 2026

Another dinobaby post. No AI unless it is an image. This dinobaby is not Grandma Moses, just Grandpa Arnold.

I have seen the name “David Sachs” in the context of affluent Silicon Valley luminaries. He has or is working in an important job in the US government. I scanned a “tweet” or is it a “musk” on X.com. The lingo in the message is not dinobaby speak, so I have to interpret it the best I can.

The main idea is that an outfit named Anthropic, the marketing juggernaut that can make paradoxes as familiar as the smell of eggs on the grill in the Palo Alto Creamery. Anthropic is into “safe” and it releases Mythos to make software unsafe and then releases a public version that is sort of safe. The folks who care about unsafe are the accelerationists and the bad actors. Most people don’t care, understand, or want to know. But BAITs (big AI tech firms) and BAIT surfers care and care a lot.

Thanks, MidJourney. Good enough.

This Anthropic stuff is disrupting the old stand by BAIT outfits.

What’s the fix? I think a hint of how the AI world works is revealed in the cited tweet or musk. I quote the post:

Fable is Mythos with guardrails. But if those guardrails fail, then you’ve exposed Mythos and its advanced cyber capabilities to people who shouldn’t have them. (Keep in mind that Anthropic itself widely promoted the idea that Mythos was a cyberweapon and needed to be regulated as such. They asked for government regulation of Mythos and championed the guardrails on Fable. If there is a vulnerability — big or small — it is Anthropic’s responsibility to patch.)

I think this means that Anthropic is not behaving like a responsible BAIT. Therefore, Anthropic shape or face the wrath of the other BAITers.

The tweet or musk continues:

A highly credible trusted partner of both Anthropic and the USG who was testing Fable came forward with a jailbreak of those guardrails. The Admin asked Dario to fix the jailbreak or de-deploy the model. Dario refused.

One of my team told me that the “trusted partner” was Amazon, an outfit I once described as the Bezos bulldozer. (Hey, sorry about those mom and pop businesses. Vrrooom vrrooom.) I know there are not too many BAIT outfits operating in the US so I am probably incorrect. Plug in your own estimable outfit.

The consequence of the “alert” was, and I quote:

In reaction, the Admin issued the export control. The Admin did this reluctantly. It’s been very surprised that Anthropic hasn’t wanted to cooperate with a reasonable safety request (ie fixing the jailbreak issue). Anthropic’s reaction is very much at odds with their branding and ethos as a safe AI research community.

“Admin” means, I think, the US Federal government. I think Anthropic was in hot water over some confusion about the normal behavior of senior government professionals. Many of these people assume that “we pay, vendors obey.” Anthropic did not get the message about this standard operating procedure.

Several observations manifested themselves on my writing pad as I was thinking about this tweet or musk. As a dinobaby, I shall exercise my prerogative and insert them in this blog;

1. US government policy can result when a single BAIT tattles on another BAIT. Both BAITs have business entanglements. (See, for instance, “Anthropic and Amazon Expand Collaboration for Up to 5 Gigawatts of New Compute.”) I wonder what the phrase “trusted partner” means to these outfits?
2. Anthropic seems to be like the fur on a Rhodesian ridgeback. The BAIT fur is tidy, fluid, and goes with the grain. The fur on the Anthropic dog goes in a different direction. The nail that sticks up gets pounded down. Does fur work that way in BAIT Land?
3. Some of the Mythos functionality can be replicated in other BAIT systems. One of my team suggested that Qwen has some utility in juicing the activities of bad actors. I think this means that a Kentucky Derby winner’s barn has burned as the valuable steed has headed off to another farm on the Bourbon Trail.

Net net:  If the tweet or the musk are accurate, policy is indeed set by what some individuals call :

• Squealing
• Dry snitching
• Grassing
• Tattling
• Ratting
• Spilling tea (organic green, of course).

Stephen E Arnold, June 25, 2026

Google isn’t the only company investing money in skilled laborers. Meta and Monday are doing similar investments. With the $50 million, Google plans to support fourteen labor unions and reach 300,000 workers across twenty states. The plan is to better educate new workers and level up training/outreach. Unlike its counterparts who want the new wave of workers to construct data centers, Google simply wants people interested in trades and will support them “no matter where they want to work.”

Here’s how Google plans to work with the trades:

“TradesFutures, an organization created by North America’s Building Trades Unions, will increase union construction career access by scaling placement for apprenticeship readiness and registered apprenticeship programs. Google previously had invested in the electrical training Alliance — an organization created by the International Brotherhood of Electrical Workers and the National Electrical Contractors Association — intending to train 100,000 electrical workers and 30,000 new apprentices in the U.S. Now, etA will aid local training programs by bringing resources to infrastructure hubs in a mobile training center pilot. Meanwhile, the United Association’s International Training Fund, in partnership with the Mechanical Contractors Association of America, will develop a five-year roadmap for scaling the plumbing, heating, ventilation, air conditioning and refrigeration sectors. That will help give workers skills necessary for the construction and maintenance of infrastructure for data center builds. Finally, the International Training Institute for the Sheet Metal and Air Conditioning Industry, backed by the Sheet Metal, Air, Rail, and Transportation Workers and Sheet Metal and Air Conditioning Contractors, will modernize its coursework and apprentice support and implement new AI tools.”

Skilled trade workers are an important part of any Googley economy. I wonder if Google knows that some of its employees as well as assorted art history majors will be lining up to be trained. Training works if the person being trained tunes in.

Whitney Grace, June 24, 2026

Another dinobaby post. No AI unless it is an image. This dinobaby is not Grandma Moses, just Grandpa Arnold.

One of the more interesting Israeli intelware service firms is BlackCore. I know most people have not heard of the company and the firm does not exactly put up billboards on Route 101 from San Francisco to the Plastic Fantastic of Silicon Valley. However, information from France about a firm that does the Cambridge Analytica social media manipulation thing has been reported in stories from Russia to Yahoo News (“Secrets of BlackCore: Israel’s Information Warfare Now Has Few Limits or Boundaries”). In the US the news appears to pivot on sports, and that is one interesting characteristic of knowledge flow in the US of A today.

Thanks, Venice.ai. You did not hang, time out, or tell me my image prompt violated your perception of proper generated imagery. Quite a surprise!

BlackCore, also rumored to do business as Mycelium, uses a range of methods to shape or weaponized certain information. The approach is standard operating procedure for lots of companies, independent contractors, and PR outfits. Therefore, it is difficult for me to get excited about this flurry of news stories. BlackCore and its inauthentic information angle may have blipped French radar during recent elections. But I am in rural Kentucky and relying on my experience about what catches the eye of French intelligence professionals. In general, it is not a great idea to light up their radar. Things happen. Case in point: The flood of articles outing the hard working team at BlackCore and its affiliates.

I would point out that some of the information may have originated intentionally or by “accident” from an intel outfit in the French government. One of my team told me that the original chunk of information about BlackCore evolved from a unit possibly affiliated with Secretariat-General for Defense and National Security (SGDSN). This outfit reports directly to the French prime minister.

Several questions occur to me:

1. Why is Meta fixated on NSO Group? That is, in my opinion, a precursor to more sophisticated intelware operations. Perhaps Meta does not know who is doing interesting things on Meta. The Zuck just falls back on his perception about the Pegasus crowd.
2. What sparked the flurry of news stories about an outfit that has done a pretty good job of remaining off the grid occupied by journalists, do-gooders, and civilians? Why now?
3. What will be done about BlackCore and similar entities operating in other countries; for example, China and Russia? My thought is that not much has been done and not much will be done. The fact that Cambridge Analytica left a game plan for every misinformation operation worldwide is one of its stellar contributions. Why was it allowed to operate with considerable freedom for several years?

Net net: Snappy name aside, BlackCore is just one of a flock of similar intelware operations. A bit of poking around on LinkedIn-type services provide plenty of hints about who is doing what.

PS. Last time I was in France I tried to get a hat with the entity name “Viginum” on it. I failed.

Stephen E Arnold, June 23, 2026

“ ‘Children will be given back their childhoods,’ the government said this morning as it unveiled more detail and stated an aim to ‘capture user-to-user platforms, whose purpose is to enable social interaction and?which allow users to post material, alongside algorithms.’”

Nothing is better than speedy action when it comes to social media. It’s been about 20 years and now that evidence of numerous reading, writing, and arithmetic type knowledge gaps, bans are coming. Norway just put down its snowshoe clad foot too.

But the UK is going further than Australia and is claiming “world leading action” related to banning social media.? ?For example, romantic-type chatbots will be required to enforce a minimum age of eighteen for usage. The government also wants to break the doom scrolling addiction among younger social media users. Personally I am not sure that the kids will be supportive of that effort.

Ofcom, a UK organizations that has the power to enforce online safety regulations, will conduct a study on which age is the most effective way to verify a person is over sixteen. A government survey also found that 90% of adults and large portion of kids are favor of the ban.

This might sour relations with the US:

“The Australian social media ban has been controversial, with American tech giants unsurprisingly reacting with alarm. As these giants get closer and closer to Donald Trump in the States, Starmer, who has consistently attempted to keep relations with the U.S. on a solid footing, may find himself falling foul of the U.S. President, although the POTUS is at present distracted with Iran.”

It’s probably going to be as effective as the media ban in Australia. In other words, they’re twenty years too late. Good luck Starmer, you’re already doing a bang up job protecting kids, especially girls. Gee, are females negatively affected by social media?

Whitney Grace, June 23, 2026