How Does One Pay for AI? Maybe Cut Prices and Make Money on Volume? (I Have a Bridge to Sell You Cheap)
May 21, 2025
Just the dinobaby operating without Copilot or its ilk.
Mucho AI marketing from the Google and the Softies. Let’s shift gears and talk about discounts similar to Filene’s Basement sale. A change of pace may put the marketing drag racers in context.
Bloomberg, the terminal people who given Thomson Reuters headaches, published “Salesforce Cuts Slack Price for US Government, Following Google.” The write up explains that lower prices for government customers is now in effect. In my experience, price cutting to get US government sales often leads to some issues. The normal mechanisms involve responding to work in Congress related to appropriations for the coming fiscal year; that is, new money for 2026-2027, one-to-one interaction to move a problem to a Request for Information and then to a project to assist in formulating and writing a Statement of Work, putting bloodhounds on the trail of end-of-year unspent funds, and a couple of other methods.
Price cutting? Well, perhaps if certain conditions have been met. I don’t want to go into these, but you can ask around for individuals who have not had their career path altered with the special deal, lower prices, and annexes to cover what are often inevitable problems with the products or service given a price cut.
Why is this important? For most people, selling services to the US government is handled through specific methods. Fancy dancing is a topic for a luncheon meeting in some organizations, but it is not as popular as talking about Kentucky basketball or the new distillery on the Bourbon Trail.
I find it interesting that Google is cutting prices for the US government. I am not sure what Amazon is doing. There was a burst of activity several years ago, but now the chatter is Microsoft, Microsoft’s deal with Palantir, and Microsoft’s security posture. Google and Salesforce? Sure, maybe.
My concern with price cuts is that Google and Salesforce are infusing smart software into their products and services. Therefore, the investments in said smart technology have to return a profit. How does one return a profit with US government sales by cutting prices? Mind you, those cuts are coming as the pressure on firms to generate a return on their investments in smart software is looking like it is exponentiating. Imagine: Exponentiating text messages, emails, and maybe face-to-face meetings in actual physical conference rooms going up every day or so. Those automated calendars are not a pretty sight in my opinion.
Several observations:
- Price cuts. Hmmm.
- Get more government customers with a K-Mart blue light special. Hmmm
- Assurances of timely service. Hmmm.
Net net: Hmmm. Discounts. Okay.
Stephen E Arnold, May 21, 2025
Yo, Open Source Cheerleaders: Department of Defense News
May 21, 2025
Add this to the many changes we have recently seen in the federal government: We learn from Tech Radar, “Pentagon Looks to Shake Up ‘Outdated’ Software Procurement, Declares War on Open Source.” As much as we love open-source software, we know it poses certain security risks for sensitive systems. With an initiative dubbed the Software Fast-Track (SWFT), DOD CIO Katherine Arrington aims to overhaul the department’s software acquisition, authorization, and testing processes. The new framework is to be published by the end of July. Writer Craig Hale reports:
“In the memo, Arrington explained the SWFT Framework will define ‘clear’ and ‘specific’ cybersecurity and Supple Chain Risk Management (SCRM) requirements, rigorous software security verification processes, secure information sharing mechanisms and Federal Government-led risk determinations to expedite the cybersecurity authorizations for rapid software adoption. She continued to explain that current systems are best seen as ‘outdated,’ noting that acquisition processes don’t enable the agility that departments need. Arrington also noted that the use of open source software ‘presents a significant and ongoing challenge,’ with a lack of visibility into the origins and security of software code particularly troubling. Malware and partner leaks have already exposed vulnerabilities in DOD systems, with software vulnerabilities among the most popular entry points for attackers.”
Excellent point. We note the DOD seems to have several goals for this initiative. One can only hope security will take precedence over rapid adoption and penny-pinching. We are curious to see how the agency will save money while shifting away from free software.
Cynthia Murrell, May 21, 2025
China Smart, US Dumb: Twisting the LLM Daozi
May 12, 2025
No AI, just the dinobaby expressing his opinions to Zellenials.
That hard-hitting technology information service Venture Beat published an interesting article. Its title is “Alibaba ZeroSearch Lets AI Learn to Google Itself — Slashing Training Costs by 88 Percent.” The main point of the write up, in my opinion, is that Chinese engineers have done something really “smart.” The knife at the throat of US smart software companies is cost. The money fires will flame out unless more dollars are dumped into the innovation furnaces of smart software.
The Venture Beat story makes the point that “could dramatically reduce the cost and complexity of training AI systems to search for information, eliminating the need for expensive commercial search engine APIs altogether.”
Oh, oh.
This is smart. Buring cash in pursuit of a fractional improvement is dumb, well, actually, stupid, if the write up’s inforamtion is accurate.
The Venture Beat story says:
The technique, called “ZeroSearch,” allows large language models (LLMs) to develop advanced search capabilities through a simulation approach rather than interacting with real search engines during the training process. This innovation could save companies significant API expenses while offering better control over how AI systems learn to retrieve information.
Is this a Snorkel variant hot from Stanford AI lab?
The write up does not delve into the synthetic data short cut to smart software. After some mumbo jumbo, the write up points out the meat of the “innovation”:
The cost savings are substantial. According to the researchers’ analysis, training with approximately 64,000 search queries using Google Search via SerpAPI would cost about $586.70, while using a 14B-parameter simulation LLM on four A100 GPUs costs only $70.80 — an 88% reduction.
Imagine. A dollar in cost becomes $0.12. If accurate, what should a savvy investor do? Pump money into an outfit like OpenAI or the Xai- type entity, or think harder about the China-smart solution?
Venture Beat explains the implication of the alleged cost savings:
The impact could be substantial for the AI industry.
No kidding?
The Venture Beat analysts add this observation:
The irony is clear: in teaching AI to search without search engines, Alibaba may have created a technology that makes traditional search engines less necessary for AI development. As these systems become more self-sufficient, the technology landscape could look very different in just a few years.
Yep, irony. Free transformer technology. Free Snorkle technology. Free kinetic into the core of the LLM money furnace.
If true, the implications are easy to outline. If bogus, the China Smart, US Dumb trope still captured ink and will be embedded in some smart software’s increasingly frequent hallucinatory outputs. At which point, the China Smart, US Dumb information gains traction and becomes “fact” to some.
Stephen E Arnold, May 12, 2025
European Union: Academics and Researchers, Come On Over to Freedom
May 8, 2025
No AI, just the dinobaby expressing his opinions to Zellenials.
Is the European Union actively advertising employment opportunities in Western Europe? Do canines sniff? I think the answer is, “Yes.”
I spotted an official European Commission announcement with the title or one of the titles: “Choose Europe. Advance Your Research Career in the EU.” Another title in the official online statement is, “Choose Groundbreaking Research.” The document says,
As a world-leading centre for research and innovation with freedom of science, the European Union offers an ideal environment to advance your career. With a wealth of stable and predictable funding opportunities and cutting-edge facilities, the EU enables researchers to work on projects where they can truly make a difference. You’ll join a dynamic and international community of top talent, all dedicated to finding solutions to the world’s biggest challenges. Europe offers an excellent quality of life, including affordable healthcare and education, excellent working conditions and strong social security for you and your family. You’ll also enjoy freedoms and protections based on our values.
I found the word choice quite interesting; for example:
affordable healthcare and education
dynamic and international community
excellent working conditions
freedom and freedoms (bang, bang!)
ideal environment
protections
quality of life
strong social security
top talent
values
world-leading
The word choice reveals what the EU thinks will appeal to some American academics and researchers as well as to others in different countries. One might think that this employment advertisement is identifying specific issues associated with certain non-EU countries.
To make the opportunity more concrete, the write up presents these data:
If I were young again, this type of lingo might appeal to me. I, however, am a dinobaby. Becoming a big-time academic researcher is a non-starter for me. For some, however, the EU’s inducement might be compelling. I have done projects and spent a reasonable amount of time in London and Paris. My son attended two universities in France, and I am not sure he wanted to return to the US, but the French government had other ideas for a 20 something.
Interesting. Opportunity with a possible message for some working in less salubrious situations. Crafty message and straight ahead marketing.
Stephen E Arnold, May 8, 2025
US Brain Drain Droplet May Presage a Beefier Outflow
May 8, 2025
Believe it or not, no smart software. Just a dumb and skeptical dinobaby.
When I was working on my PhD at the University of Illinois, I noticed that the number of foreign students on campus seemed to go up each year. One year in the luxurious Florida Avenue Residence Hall, most of the students were from farms. The next year, FAR was a mini-United Nations. I did not pay any attention because I was on my way to an actual “real” job at Halliburton Nuclear in Washington, DC.
I heard the phrase “brain drain” over the years. The idea was that people who wanted to work in technical fields would come to the US, get degrees, and then stay to work in US universities or dolphin-loving, humanity-centric outfits like the nuclear industry. The idea was that the US was a magnet: Good schools, many opportunities to work or start a company.
I am not sure that golden age exists any longer. I read about universities becoming research labs for giant companies. I see podcasts with foaming-at-the-mouth academics complaining about [a] the quality of the students, [b] squabbles between different ideological groups, and [c] the lack of tenure opportunities which once seemed to be a sinecure for life just like the US government’s senior executive service.
Now the world works in ever more mysterious ways. As a confused dinobaby, I read news items (unverified, of course) with headlines like this:
Top US Scientist leaves Department Of Energy To Join Sichuan University Amid Rising China Tensions.
The write up reports a “real” news:
Amid escalating US-China tensions, senior scientist Yi Shouliang, formerly with the US Department of Energy, has left the U.S. to assume a new academic role at Sichuan University in China…. Shouliang served as a principal scientist and project leader at the DOE’s National Energy Technology Laboratory (NETL), where he focused on the Water-Energy Program.
Let’s assume that this academic who had some business interests just missed his family. No big deal.
But what if a certain “home” country was starting to contact certain people and explaining that their future was back in the good old homeland? Could that country systematically explain the facts of life in a way that made the “home” country look more appealing than a big house in Squirrel Hill?
For a few months, I have been writing “China smart, US dumb” blog posts when I spot some news about how wonderfully bright many young Chinese men and women are.
As a dinobaby, my first thought is that China wants its smart people back in the Middle Kingdom. Hopefully more information about this 2025 brain drain from the US to other countries will become publicly available. Plus, one isolated person going against the “You can’t go home again” idea means nothing. Or does it mean something is afoot?
PS. No, I never went back to Chambana to turn in my thesis. I liked working at Halliburton Nuclear more than I liked indexing poetry for the now departed Dr. William Gillis. Sorry, Dr. Gillis, the truth is now out.
Stephen E Arnold, May 8, 2025
China Tough. US Weak: A Variation of the China Smart. US Dumb Campaign
May 6, 2025
No AI. This old dinobaby just plods along, delighted he is old and this craziness will soon be left behind. What about you?
Even members of my own team thing I am confusing information about China’s technology with my dinobaby peculiarities. That may be. Nevertheless, I want to document the story “The Ancient Chinese General Whose Calm During Surgery Is Still Told of Today.” I know it is. I just read a modern retelling of the tale in the South China Morning Post. (Hey. Where did that paywall go?)
The basic idea is that a Chinese leader (tough by genetics and mental discipline) had dinner with some colleagues. A physician showed up and told the general, “You have poison in your arm bone.”
The leader allegedly told the physician,
“No big deal. Do the surgery here at the dinner table.”
The leader let the doc chop open his arm, remove the diseased area, and stitched the leader up. Now here’s the item in the write up I find interesting because it makes clear [a] the leader’s indifference to his colleagues who might find this surgical procedure an appetite killer and [b] the flawed collection of blood which seeped after the incision was made. Keep in mind that the leader did not need any soporific, and the leader continued to chit chat with his colleagues. I assume the leader’s anecdotes and social skills kept his guests mesmerized.
Here’s the detail from the China Tough. US Weak write up:
“Guan Yu [the tough leader] calmly extended his arm for the doctor to proceed. At the time, he was sitting with fellow generals, eating and drinking together. As the doctor cut into his arm, blood flowed profusely, overflowing the basin meant to catch it. Yet Guan Yu continued to eat meat, drink wine, and chat and laugh as if nothing was happening.”
Yep, blood flowed profusely. Just the extra that sets one meal apart from another. The closest approximation in my experience was arriving at a fast food restaurant after a shooting. Quite a mess and the odor did not make me think of a cheeseburger with ketchup.
I expect that members of my team will complain about this blog post. That’s okay. I am a dinobaby, but I think this variation on the China Smart. US Dumb information flow is interesting. Okay, anyone want to pop over for fried squirrel. We can skin, gut, and fry them at one go. My mouth is watering at the thought. If we are lucky, one of the group will have bagged a deer. Now that’s an opportunity to add some of that hoist, skin, cut, and grill to the evening meal. Guan Yu, the tough Chinese leader, would definitely get with the kitchen work.
Stephen E Arnold, May 6, 2025
AI Chatbots Now Learning Russian Propaganda
May 6, 2025
Gee, who would have guessed? Forbes reports, “Russian Propaganda Has Now Infected Western AI Chatbots—New Study.” Contributor Tor Constantino cites a recent NewsGuard report as he writes:
“A Moscow-based disinformation network known as ‘Pravda’ — the Russian word for ‘truth’ — has been flooding search results and web crawlers with pro-Kremlin falsehoods, causing AI systems to regurgitate misleading narratives. The Pravda network, which published 3.6 million articles in 2024 alone, is leveraging artificial intelligence to amplify Moscow’s influence at an unprecedented scale. The audit revealed that 10 leading AI chatbots repeated false narratives pushed by Pravda 33% of the time. Shockingly, seven of these chatbots directly cited Pravda sites as legitimate sources. In an email exchange, NewsGuard analyst Isis Blachez wrote that the study does not ‘name names’ of the AI systems most susceptible to the falsehood flow but acknowledged that the threat is widespread.”
Blachez believes a shift is underway from Russian operatives directly targeting readers to manipulation of AI models. Much more efficient. And sneaky. We learn:
“One of the most alarming practices uncovered is what NewsGuard refers to as ‘LLM grooming.’ This tactic is described as the deliberate deception of datasets that AI models — such as ChatGPT, Claude, Gemini, Grok 3, Perplexity and others — train on by flooding them with disinformation. Blachez noted that this propaganda pile-on is designed to bias AI outputs to align with pro-Russian perspectives. Pravda’s approach is methodical, relying on a sprawling network of 150 websites publishing in dozens of languages across 49 countries.”
AI firms can try to block propaganda sites from their models’ curriculum, but the operation is so large and elaborate it may be impossible. And also, how would they know if they had managed to do so? Nevertheless, Blachez encourages them to try. Otherwise, tech firms are destined to become conduits for the Kremlin’s agenda, she warns.
Of course, the rest of us have a responsibility here as well. We can and should double check information served up by AI. NewsGuard suggests its own Misinformation Fingerprints, a catalog of provably false claims it has found online. Or here is an idea: maybe do not turn to AI for information in the first place. After all, the tools are notoriously unreliable. And that is before Russian operatives get involved.
Cynthia Murrell, May 6, 2025
The EU Bumps Heads with Tech Bros
May 1, 2025
Dinobaby, here. No smart software involved unlike some outfits. I did use Sam AI-Man’s art system to produce the illustration in the blog post.
I noticed some faint signals that the European Union has bumped heads with a couple of US tech bros. The tech bros have money, users, and a do-it-my way attitude. The EU moves less quickly and likes to discuss lunch before going to lunch. The speedy delivery approach upsets stomachs of some European professionals.
The soccer player on a team sponsored by tech bros knocks over the old player and wins the ball. The problem is that the youthful, handsome, well-paid superstar gets a red card. Thanks, OpenAI. I am looking forward to your Telegram clone.
The hints of trouble appear in “Brussels Takes Action Against Google and Apple Despite Trump Threat.” The article explains that the tech bros have violated the Digital Markets Act. Some pundits have suggested that the DMA exists because of certain tech bros and their zip-zip approach to shaping monopolistic business methods.
What are the US tech bros going to do? [a] Posture, [b] output PR, [c] litigate, [d] absolutely everything possible. The answer, based on my limited understanding of how big time thinkers with money and win-at-all-costs logic business executives thing, [d].
Let’s think about how this disagreement will unfold.
First, the use of media to communicate the unfairness of a governmental entity telling a couple of tech bros they can’t race their high performance vehicles down Avenue Louise, the Ku’damm, or the Champs-Élysées. Then the outfits will output PR, lots of PR. Third, the lawyers will take flight. If there are not enough legal eagles in Europe, convocations will be whisked to Brussels and Strasbourg. The final step will open the barn door and let the animals run free.
With the diplomatic skills of a SWAT team and piles of money, the afflicted tech bros will try to get the EU to knock off the anti-tech-bro double talk. Roll over or ….?
That’s the question, “Or what?”
The afflicted tech bros are accustomed to doing what they want, using slick talk and other inducements to do exactly what they want. The “you want to move the icons on the home screen” and “you want objective search results” attitudes are likely to be somewhat ineffective.
I am not sure what the tech bros will do. France broke the wing of Telegram’s big bird. After realizing that France could put him in a depressingly over crowded prison about 16 kilometers from a five star hotel in Paris, the Telegram tech bro complied.
Will the defendants in future legal disputes with the EU show up in court to explain to the slug-like thinkers in the EU government bureau that they must do what the US tech bros want. There’s that “or” again. It is a pesky matter.
Tech bros, as Pavel Durov learned in his seven months of intensive classes in French law, the bureaucracy moves slowly and has a variety of financial levers and knobs. These can be adjusted in numerous ways. It is indeed possible that if a tech bro gets out of line, he could experience a crash course in EU systems and procedures.
The inconceivable could happen: The companies products could be constrained in some way. With each “do it our way” output, the knobs and dials can be adjusted.
Could the tie up of Ecosia and Wolfram Alpha or Swisscows offer a viable option for search? Could the Huawei-type of mobile devices replace the iPhone?
The tech bros may want to check out how Pavel Durov’s approach to business is working out.
Stephen E Arnold, May 1, 2025
France And Germany Form Open Source Writing Collaboration
April 30, 2025
Open source software and AI algorithms are a match made in heaven. You can’t say the same thing about France and Germany when it comes to history, but the countries can put aside their differences (occasionally) to advance technology. The French and German governments came together to design Docs.
Docs is described as “Collaborative writing, simplified-collaborate and write in real time, without layout constraints.” I don’t know if the term “layout” refers to a writing software’s formatting or if it means limited to the constraints of writing software. It could mean either of things or something is lost in the literal translation. Ich habe keine Ahnung. Je ne c’est pas.
Docs is built on the Django Rest Framework and Nest.js. It also uses BlockNote.js and Yes (they also sponsor those text editors too). Docs can be self-hosted, has a business friendly license, and welcomes anyone to contribute to its growth either monetarily or via code). Here is what Docs offers as a writing partner:
“Docs offers an intuitive writing experience. Its minimalist interface favors content over layout, while offering the essentials: media import, offline mode and keyboard shortcuts for greater efficiency.”
So far that sounds très magnifique and ausgezeichnet! Docs also offers simple real-time collaboration. Users on a document can access the same document, see changes made live, and maintain control of the document for data security. Docs also has universal formats for exportation: OpenDocument, Word, and PDF.
A nifty feature unavailable with most writing software is the ability to organize documents into knowledge bases with subpages. This feature also comes with search and pinning capabilities.
This French and German writing collaboration sounds amazing! Break out the champagne and beer and enjoy some croissants and pretzels. This is one open source tool everyone needs!
Whitney Grace, April 30, 2025
Apple and Meta: Virtual Automatic Teller Machines for the EU
April 29, 2025
No AI, just a dinobaby watching the world respond to the tech bros.
I spotted this story in USA Today. You remember that newspaper, of course. The story “Apple Fined $570 Million and Meta $228 Million for Breaching European Union Law” reports:
Apple was fined 500 million euros ($570 million) on Wednesday and Meta 200 million euros, as European Union antitrust regulators handed out the first sanctions under landmark legislation aimed at curbing the power of Big Tech.
I have observed that to many regulators the brands Apple and Meta (Facebook) are converted to the sound of ka-ching. For those who don’t recognize the onomatopoeia for an old-fashioned cash register ringing up a sale. The modern metaphor might be an automatic teller machine emitting beeps and honks. That works. Punch the Apple and Meta logos and bonk, beep, out comes millions of euros. Bonk, beep.
The law which allows the behavior of what some Europeans view as “tech bros” to be converted first to a legal process and then to cash is the Digital Markets Act. The idea is that certain technology centric outfits based in the US operate without much regard for the rules, regulations, and laws of actual nation-states and their governing entities. I mean who pays attention to what the European Union says? Certainly not a geek à la sauce californienne.
The companies are likely to interpret these fines as some sort of deus ex machina, delivered by a third-rate vengeful god in a TikTok-type of video. Perhaps? But the legal process identified some actions by the fined American companies as illegal. Examples range from preventing an Apple store user from certain behaviors to Meta’s reluctance to conform to some privacy requirements. I am certainly not a lawyer, nor am I involved with either of the American companies. However, I can make several observations from my dinobaby point of view, of course:
- The ka-ching / bonk beep incentive is strong. Money talks in the US and elsewhere. It is not surprising that the fines are becoming larger with each go-round. How does one stop the cost creep? One thought is to change the behavior of the companies. Sorry, EU, that is not going to happen.
- The interpretation of the penalty as a reaction against America is definitely a factor. For those who have not lived and worked in other countries, the anti-American sentiment is not understood. I learned when people painted slurs on the walls of our home in Campinas, Brazil. I was about 13, and the anger extended beyond black paint on our pristine white, eight-foot high walls with glass embedded at the top of them. Inviting, right?
- The perception that a company is more powerful than a mere government entity has been growing as the concentration of eyeballs, money, and talented people has increased at certain firms. Once the regulators have worked through the others in this category, attention will turn to the second tier of companies. I won’t identify any entities but the increased scrutiny of Cloudflare by French authorities is a glimpse of what might be coming down the information highway.
Net net: Ka-ching, ka-ching, and ka-ching. Beep, bong, beep, bong.
Stephen E Arnold, April 29, 2025
-
- Subscribe to Beyond Search
Feature archive
News archive
-
