At Google Innovation Never Stops or Gee a G
October 10, 2025
This essay is the work of a dumb dinobaby. No smart software required.
I read “Google’s Gradient G Icon Design Is Going Company Wide.” Usually Deepseek, the YouTube leadership, or a rando in advertising announces a quantumly supreme achievement. The stunning Google news for September 29, 2025, is presented this way:
Google used “brighter hues and gradient design” to “symbolize the surge of AI-driven innovation and creative energy across our products and technology.” The aim was to stay “true to Google’s iconic four colors,” with the last design refresh taking place 10 years ago.
The article includes the old G and the new forward leaning, innovative, quantumly supreme G. Here’s what I saw in the cited write up:
This is the old, backward leaning, non-innovative, un-quantumly supreme G.
Now here’s is the new forward leaning, innovative, quantumly supreme G:
That is revolutionary, boundary stretching, Leonardo DaVinci grade art.
I am impressed. Imagine the achievement amidst some staff concern about layoffs, and the financial headaches resulting from those data center initiatives, crypto services, and advertising sales efforts.
What’s next from the Google? Gee, this new G will be difficult to galvanize more grandiose game changers.
Stephen E Arnold, October 10, 2025
ChatGPT Finds Humans Useful
October 10, 2025
OpenAI is chasing consumers during primetime football games, we learn from 9to5Mac’s piece, “Pressure Mounts on Siri as ChatGPT Ads Start Airing on Primetime TV.” The first of these ads premiered during NFL Primetime. We are told the campaign focuses on ways people are using ChatGPT in their everyday lives, like creating recipes or fitness plans. So wholesome! (We assume they are leaving out the many downsides of overreliance on the tech.) Does this mean firm’s second Super Bowl ad will be more down to earth than its first one?
Writer Ben Lovejoy asserts this campaign highlights how embarrassingly far Apple’s Siri is behind ChatGPT. iPhone users have the option to get an answer from ChatGPT when Siri fails them. But, as Lovejoy notes, the permission prompt serves as a spotlight on Siri’s inadequacies.
The ad campaign comes with an interesting caveat. We learn:
“With growing concern in the creative sector around the use of AI, the company has gone out of its way to ensure that no artificial intelligence was used for the actual creative work. Creative Review reports: Crucially, the campaign was created largely through human endeavour, with the team at OpenAI noting that: ‘Human craft was central to the campaign’s creation. Every frame was shot on film, shaped by directors, photographers, producers and many more masters of craft.’ That ‘largely’ rider reflects that ChatGPT was used for some background work, with ‘streamlining shot lists and organising schedules’ given as examples.”
Will this acknowledgement that real life is better than AI fakery backfire on the premier AI company? And no Sora?
Cynthia Murrell, October 10, 2025
AI Has a Secret: Humans Do the Work
October 10, 2025
A key component of artificial intelligence output is not artificial at all. The Guardian reveals “How Thousands of ‘Overworked, Underpaid’ Humans Train Google’s AI to Seem Smart.” From accuracy to content moderation, Google Gemini and other AI models rely on a host of humans employed by third-party contractors. Humans whose jobs get harder and harder as they are pressured to churn through the work faster and faster. Gee, what could go wrong?
Reporter Varsha Bansal relates:
“Each new model release comes with the promise of higher accuracy, which means that for each version, these AI raters are working hard to check if the model responses are safe for the user. Thousands of humans lend their intelligence to teach chatbots the right responses across domains as varied as medicine, architecture and astrophysics, correcting mistakes and steering away from harmful outputs.”
Very important work—which is why companies treat these folks as valued assets. Just kidding. We learn:
“Despite their significant contributions to these AI models, which would perhaps hallucinate if not for these quality control editors, these workers feel hidden. ‘AI isn’t magic; it’s a pyramid scheme of human labor,’ said Adio Dinika, a researcher at the Distributed AI Research Institute based in Bremen, Germany. ‘These raters are the middle rung: invisible, essential and expendable.’”
And, increasingly, rushed. The write-up continues:
“[One rater’s] timer of 30 minutes for each task shrank to 15 – which meant reading, fact-checking and rating approximately 500 words per response, sometimes more. The tightening constraints made her question the quality of her work and, by extension, the reliability of the AI. In May 2023, a contract worker for Appen submitted a letter to the US Congress that the pace imposed on him and others would make Google Bard, Gemini’s predecessor, a ‘faulty’ and ‘dangerous’ product.”
And that is how we get AI advice like using glue on pizza or adding rocks to one’s diet. After those actual suggestions went out, Google focused on quality over quantity. Briefly. But, according to workers, it was not long before they were again told to emphasize speed over accuracy. For example, last December, Google announced raters could no longer skip prompts on topics they knew little about. Think workers with no medical expertise reviewing health advice. Not great. Furthermore, guardrails around harmful content were perforated with new loopholes. Bansal quotes Rachael Sawyer, a rater employed by Gemini contractor GlobalLogic:
“It used to be that the model could not say racial slurs whatsoever. In February, that changed, and now, as long as the user uses a racial slur, the model can repeat it, but it can’t generate it. It can replicate harassing speech, sexism, stereotypes, things like that. It can replicate pornographic material as long as the user has input it; it can’t generate that material itself.”
Lovely. It is policies like this that leave many workers very uncomfortable with the software they are helping to produce. In fact, most say they avoid using LLMs and actively discourage friends and family from doing so.
On top of the disillusionment, pressure to perform full tilt, and low pay, raters also face job insecurity. We learn GlobalLogic has been rolling out layoffs since the beginning of the year. The article concludes with this quote from Sawyer:
‘I just want people to know that AI is being sold as this tech magic – that’s why there’s a little sparkle symbol next to an AI response,’ said Sawyer. ‘But it’s not. It’s built on the backs of overworked, underpaid human beings.’
We wish we could say we are surprised.
Cynthia Murrell, October 10, 2025
AI Embraces the Ethos of Enterprise Search
October 9, 2025
This essay is the work of a dumb dinobaby. No smart software required.
In my files, I have examples of the marketing collateral generated by enterprise search vendors. I have some clippings from trade publications and other odds and ends dumped into my enterprise search folder. One of these reports is “Fastgründer John Markus Lervik dømt til fengsel.” The article is no longer online, but you can read my 2014 summary at this Beyond Search link. The write up documents an enterprise search vendor who used some alleged accounting methods to put a shine on the company. In 2008, Microsoft purchased Fast Search & Transfer putting an end to this interesting company.
A young CPA MBA BA (with honors) is jockeying a spreadsheet. His father worked for an enterprise search vendor based in the UK. His son is using his father’s template but cannot get the numbers to show positive cash flows across six quarters. Thanks, Venice.ai. Good enough.
Why am I mentioning Fast Search & Transfer? The information in Fortune Magazine’s “‘There’s So Much Pressure to Be the Company That Went from Zero to $100 Million in X Days’: Inside the Sketchy World of ARR and Inflated AI Startup Accounting” jogged my memory about Fast Search and a couple of other interesting companies in the enterprise search sector.
Enterprise search was the alleged technology to put an organization’s information at the fingertips of employees. Enterprise search would unify silos of information. Enterprise search would unlock the value of an organization’s “hidden” or “dark” data. Enterprise search would put those hours wasted looking for information to better use. (IDC was the cheerleader for the efficiency payoff from enterprise search.)
Does this sound familiar? It should every vendor applying AI to an organization’s information challenges is either recycling old chestnuts from the Golden Age of Enterprise Search or wandering in the data orchard discovering these glittering generalities amidst nuggets of high value jargon.
The Fortune article states:
There’s now a massive amount of pressure on AI-focused founders, at earlier stages than ever before: If you’re not generating revenue immediately, what are you even doing? Founders—in an effort to keep up with the Joneses—are counting all sorts of things as “long-term revenue” that are, to be blunt, nothing your Accounting 101 professor would recognize as legitimate. Exacerbating the pressure is the fact that more VCs than ever are trying to funnel capital into possible winners, at a time where there’s no certainty about what evaluating success or traction even looks like.
Would AI start ups fudge numbers? Of course not. Someone at the start up or investment firm took a class in business ethics. (The pizza in those study groups was good. Great if it could be charged to another group member’s Visa without her knowledge. Ho ho ho.)
The write up purses the idea that ARR or annual recurring revenue is a metric that may not reflect the health of an AI business. No kidding? When an outfit has zero revenue resulting from dumping investor case into a burning dumpster fire, it is difficult for me to understand how people see a payoff from AI. The “payoff” comes from moving money around, not from getting cash from people or organizations on a consistent basis. Subscription-like business models are great until churn becomes a factor.
The real point of the write up for me is that financial tricks, not customers paying for the product or service, are the name of the game. One big enterprise search outfit used “circular” deals to boost revenue. I did some small work for this outfit, so I cannot identify it. The same method is now part of the AI revolution involving Nvidia, OpenAI, and a number of other outfits. Whose money is moving? Who gets it? What’s the payoff? These are questions not addressed in depth in the information to which I have access?
I think financial intermediaries are the folks taking home the money. Some vendors may get paid like masters of black art accounting. But investor payoff? I am not so sure. For me the good old days of enterprise search are back again, just with bigger numbers and more impactful financial consequences.
As an aside, the Fortune article uses the word “shit” twice. Freudian slip or a change in editorial standards at Fortune? That word was applied by one of my team when asked to describe the companies I profiled in the Enterprise Search Report I wrote many years ago. “Are you talking about my book or enterprise search?” I asked. My team member replied, “The enterprise search thing.”
Stephen E Arnold, October 2025
AI Security: Big Plus or Big Minus?
October 9, 2025
Agentic AI presents a new security crisis. But one firm stands ready to help you survive the threat. Cybersecurity firm Palo Alto Networks describes “Agentic AI and the Looming Board-Level Security Crisis.” Writer and CSO Haider Pasha sounds the alarm:
“In the past year, my team and I have spoken to over 3,000 of Europe’s top business leaders, and these conversations have led me to a stark conclusion: Three out of four current agentic AI projects are on track to experience significant security challenges. The hype, and resulting FOMO, around AI and agentic AI has led many organisations to run before they’ve learned to walk in this emerging space. It’s no surprise how Gartner expects agentic AI cancellations to rise through 2027 or that an MIT report shows most enterprise GenAI pilots already failing. The situation is even worse from a cybersecurity perspective, with only 6% of organizations leveraging an advanced security framework for AI, according to Stanford.
But the root issue isn’t bad code, it’s bad governance. Unless boards instill a security mindset from the outset and urgently step in to enforce governance while setting clear outcomes and embedding guardrails in agentic AI rollouts, failure is inevitable.”
The post suggests several ways to implement this security mindset from the start. For example, companies should create a council that oversees AI agents across the organization. They should also center initiatives on business goals and risks, not shiny new tech for its own sake. Finally, enforce least-privilege access policies as if the AI agent were a young intern. See the write-up for more details on these measures.
If one is overwhelmed by the thought of implementing these best practices, never fear. Palo Alto Networks just happens to have the platform to help. So go ahead and fear the future, just license the fix now.
Cynthia Murrell, October 9, 2025
Antitrust: Can Google Dodge Guilt Again?
October 9, 2025
The US Department of Justice brought an antitrust case against Google and Alphabet Inc. got away with a slap on the wrist. John Polonis via Medium shared the details and his opinion in, “Google’s Antitrust Escape And Tech’s Uncertain Future.” The Department of Justice can’t claim a victory in this case, because none of the suggestions to curtail Google’s power will be implemented.
Some restrictions were passed that ban exclusivity deals and require data sharing, but that’s all. It’s also nothing like the antitrust outcome of the Microsoft case in the 2000s. The judge behind the decision was Amit Mehta and he did want to deliver a dose of humility to Google:
“Judge Mehta also exercised humility when forcing Google to share data. Google will need to share parts of its search index with competitors, but it isn’t required to share other data related to those results (e.g., the quality of web pages). The reason for so much humility? Artificial intelligence. The judge emphasized Google’s new reality; how much harder it must fight to keep up with competitors who are seizing search queries that Google previously monopolized across smartphones and browsers.
Google can no longer use its financial clout like it did when it was the 900 pound gorilla of search. It’s amazing how much can change between the filing of an antitrust case and adjudication (generative AI didn’t even exist!).”
Google is now free to go hog wild with its AI projects without regulation. Google hasn’t lost any competitive edge, unlike Microsoft in its antitrust litigation. They’re now free to do whatever they want as well.
Polonis makes a very accurate point:
“The message is clear. Unless the government uncovers smoking gun evidence of deliberate anticompetitive intent — the kind of internal emails and memos that doomed Microsoft in the late 1990s (“cut off Netscape’s air supply”) — judges are reluctant to impose the most extreme remedies. Courts want narrow, targeted fixes that minimize unnecessary disruption. And the remedies should be directly tied to the anticompetitive conduct (which is why Judge Mehta focused so heavily on exclusivity agreements).”
Big Tech has a barrier free sandbox to experiment and conduct AI business deals. Judge Mehta’s decision has shaped society in ways we can’t predict, even AI doesn’t know the future yet. What will the US judicial process deliver in Google’s advertising legal dust up? We know Google can write checks to make problems go away. Will this work again for this estimable firm?
Whitney Grace, October 8, 2025
Google Bricks Up Its Walled Garden
October 8, 2025
Google is adding bricks to its garden wall, insisting Android-app developers must pay up or stay out. Neowin declares, “Google’s Shocking Developer Decree Struggles to Justify the Urgent Threat to F-Droid.” The new edict requires anyone developing an app for Android to register with Google, whether or not they sell through its Play Store. Registration requires paying a fee, uploading personal IDs, and agreeing to Google’s fine print.
The measure will have a large impact on alternative app stores like F-Droid. That open-source publisher, with its focus on privacy, is particularly concerned about the requirements. In fact, it would rather shutter its project than force developers to register with Google. That would mean thousands of verified apps will vanish from the Web, never to be downloaded or updated again. F-Droid suspects Google’s motives are far from pure. Writer Paul Hill tells us:
“F-Droid has questioned whether forced registration will really solve anything because lots of malware apps have been found in the Google Play Store over the years, demonstrating that corporate gatekeeping doesn’t mean users are protected. F-Droid also points out that Google already defends users against malicious third-party apps with the Play Protect services which scan and disable malware apps, regardless of their origin. While not true for all alternative app stores, F-Droid already has strong security because the apps it includes are all open source that anyone can audit, the build logs are public, and builds are reproducible. When you submit an app to F-Droid, the maintainers help set up your repository properly so that when you publish an update to your code, F-Droid’s servers manually build the executable, this prevents the addition of any malware not in the source code.”
Sounds at least as secure as the Play Store to us. So what is really going on? The write-up states:
“The F-Droid project has said that it doesn’t believe that the developer registration is motivated by security. Instead, it thinks that Google is trying to consolidate power by tightening control over a formerly open ecosystem. It said that by tying application identifiers to personal ID checks and fees, it creates a choke point that restricts competition and limits user freedom.”
F-Droid is responding with a call for regulators to scrutinize this and other Googley moves for monopolistic tendencies. It also wants safeguards for app stores that wish to protect developers’ privacy. Who will win this struggle between independent app stores and the tech giant?
Cynthia Murrell, October 8, 2025
With or Without AI: Winners Win and Losers Lose
October 8, 2025
This essay is the work of a dumb dinobaby. No smart software required.
Some outfits are just losers. That’s the message I got after reading “AI Magnifies Your Teams’ Strengths – and Weaknesses, Google Report Finds.” Keep in mind that this report — the DORA Report or DevOps Research & Assessment — is Googley. The write up makes clear that Google is not hallucinating. The outstanding company:
surveyed 5,000 software development professionals across industries and followed up with more than 100 hours of interviews. It may be one of the most comprehensive studies of AI’s changing role in software development, especially at the enterprise level.
Winners with AI win bigger. Losers with AI continue to lose. Is that sad team mascot one of Sam Altman’s AI cheerleaders. I think it is. Thanks, MidJourney. Good enough.
Obviously the study is “one of the most comprehensive”; of course, it is Google’s study!
The big finding seems to be:
… AI has moved from hype to mainstream in the enterprise software development world. Second, real advantage isn’t about the tools (or even the AI you use). It’s about building solid organizational systems. Without those systems, AI has little advantage. And third, AI is a mirror. It reflects and magnifies how well (or poorly) you already operate.
I interpret the findings of the DORA Report in an easy-to-remember way: Losers still lose even if their teams use AI. I think of this as a dominant football team. The team has the money to induce or direct events. As a result, the team has the best players. The team has the best coaches (leadership). The team has the best infrastructure. In short, when one is the best, AI makes the best better.
On the other hand, a losing team composed of losers will use AI and still lose.
I noted that the report about DORA did not include:
- Method of sample selection
- Questions asked
- Methodology for generating the numerous statistics in the write up.
What happens if one conducts a study to validate the idea that winners win and losers keep on losing? I think it sends a clear signal that a monopoly-type of outfit has a bit of an inferiority or fear-centric tactical view. Even the quantumly supreme need a marketing pick me up now and then.
Stephen E Arnold, October 8, 2025
Slopity Slopity Slop: Nice Work AI Leaders
October 8, 2025
Remember that article about academic and scientific publishers using AI to churn out pseudoscience and crap papers? Or how about that story relating to authors’ works being stolen to train AI algorithms? Did I mention they were stealing art too?
Techdirt literally has the dirt on AI creating more slop: “AI Slop Startup To Flood The Internet With Thousands Of AI Slop Podcasts, Calls Critics Of AI Slop ‘Luddites’.” AI is a helpful tool. It’s great to assist with mundane things of life or improve workflows. Automation, however, has become the newest sensation. Big Tech bigwigs and other corporate giants are using it to line their purses, while making lives worse for others.
Note this outstanding example of a startup that appears to be interested in slop:
“Case in point: a new startup named Inception Point AI is preparing to flood the internet with a thousands upon thousands of LLM-generated podcasts hosted by fake experts and influencers. The podcasts cost the startup a dollar or so to make, so even if just a few dozen folks subscribe they hope to break even…”
They’ll make the episodes for less than a dollar. Podcasting is already a saturated market, but Point AI plans to flush it with garbage. They don’t care about the ethics. It’s going to be the Temu of podcasts. It would be great if people would flock to true human-made stuff, but they probably won’t.
Another reason we’re in a knowledge swamp with crocodiles.
Whitney Grace, October 9, 2025
Google Gets the Crypto Telegram
October 7, 2025
This essay is the work of a dumb dinobaby. No smart software required.
Not too many people cared that Google cut a deal with Alibaba’s ANT financial services outfit. My view is that at some point down the information highway, the agreement will capture more attention. Today (September 27, 2025), I want to highlight another example of Google’s getting a telegram about crypto.
Finding inspiration? Yep. Thanks, Venice.ai. Good enough.
Navigate to what seems to be just another crypto mining news announcement: “Cipher Mining Signs 168 MW, 10-Year AI Hosting Agreement with Fluidstack.”
So what’s a Cipher Mining? This is a publicly traded outfit engaged in crypto mining. My understanding is that the company’s primary source of revenue is bitcoin mining. Some may disagree, pointing to its business as “owner, developer and operator of industrial-scale data centers.”
The news release says:
[Cipher Mining] announces a 10-year high-performance computing (HPC) colocation agreement with Fluidstack, a premier AI cloud platform that builds and operates HPC clusters for some of the world’s largest companies.
So what?
The news release also offers this information:
Google will backstop $1.4 billion of Fluidstack’s lease obligations to support project-related debt financing and will receive warrants to acquire approximately 24 million shares of Cipher common stock, equating to an approximately 5.4% pro forma equity ownership stake, subject to adjustment and a potential cash settlement under certain circumstances. Cipher plans to retain 100% ownership of the project and access the capital markets as necessary to fund a portion of the project.
Okay, three outfits: crypto, data centers, and billions of dollars. That’s quite an information cocktail.
Several observations:
- Like the Alibaba / ANT relationship, the move is aligned with facilitating crypto activities on a large scale
- In the best tradition of moving money, Google seems to be involved but not the big dog. I think that Google may indeed be the big dog. Puzzle pieces that fit together? Seems like it to me.
- Crypto and financial services could — note I say “could” — be the hedge against future advertising revenue potholes.
Net net: Worth watching and asking, “What’s the next Google message received from Telegram?” Does this question seem cryptic? It isn’t. Like Meta, Google is following a path trod by a certain outfit now operating in Dubai. Is the path intentional or accidental? Where Google is concerned, everything is original, AI, and quantumly supreme.
Stephen E Arnold, October 7, 2025
-
- Subscribe to Beyond Search
Feature archive
News archive
-
