It Is Official: The New York Times Says Google AI Search Is Better
May 12, 2026
Another dinobaby post. No AI unless it is an image. This dinobaby is not Grandma Moses, just Grandpa Arnold.
Yep, Google’s AI search is better. I think the statements in “Five Ways A.I. Search Beats an Old-School Google Search.” Sure, the savvy editors tweaked the sub-title to the story to make the headline less crazy. Here the add-on is:
Google’s A.I. search technology is far from perfect (don’t count on it for celebrity news), but it excels at tasks like picking out groceries and detecting scams.
This is similar to Dennis Day’s non sequiturs crafted by the Jack Benny Show writers in the late 1940s. Oh, not PERFECT, just BETTER.
The write up consists of five topical queries. The author plugs in prompts to the Google and says:
Even though the technology [Google AI Web search service] remains imperfect, I am increasingly clicking the button, labeled “AI Mode,” on Google.com to type requests and immediately finish tasks that would have required many minutes with an old-school search…. It took me some experimenting to get better results from A.I. search, and the key was to tell Google to work with a small amount of information instead of crawling the web for answers.
This means that the criteria for “better” boils down to one person’s five consumer-type queries. The results were better. Okay, now I get to ask, “Better than what?” Other than a subjective reaction to Google output for generalist queries, what makes the results better? Are the results evidence of improved precision and recall? Are the results better than the outputs from Bing, Yandex, or a metasearch system performing retrieval value add in the manner of Kagi? Are the results more accurate than data gathered by doing what I call the “dinobaby method” of contacting experts in each product/service area? Doing time-consuming and difficult work like telephoning those who are professional researchers for a consulting firm or an academic with specific knowledge of consumer query relevance?
Thanks, MidJourney. You blundered back online and forced me to use a year old model. Nice work.
You know the answers to these questions? Here’s my summary of what I think one or more NYT professionals would tell me:
Dude, you are nuts. This is a colorful, generalist piece designed to keep Google happy, our marketing team happy, our subscribers happy, and our colleagues happy. This is a quick chunk of “real news writing.” Just stick to your dinobaby fetish about Russian crypto cash transfers and leave us “real news” people alone. Sure, you were a consultant here once, but now you are really old and the fools who paid you to assist us were buzzed with newsroom coffee.
Okay, I understand. I just returned from a couple of major conference lectures delivered to people younger than me who undoubted interpreted my approach as either worthless or just plain crazy. Money laundering. Are you kidding me? (Actually, no, there are two NASDAQ listed companies engaged in this activity as I type my comments about search.)
Here’s my take on Google search. I guarantee most people will not like my point of view. As a dinobaby, I say, “Too bad.”
First, Google search is a de facto monopoly for millions of people. As a result, “search” means whatever Google outputs. What makes this dangerous is that users of computing devices continue to believe the outputs are comprehensive, accurate, up to date, and neutral (unbiased). People, including professionals at the conferences I attend, tell me they are “experts” at using Google. They are not. OSINT professionals rely on Google for much of the work. Validation and verification are not popular pass times among most Google users. Therefore, assertions about Google are essentially valueless to a person like me. How do I know? For starters, I have written three monographs about Google for a defunct publisher called Infonortics Ltd. in the UK, and I did a column for Knowledge Management mostly about search and the Google for a number of years. I have a couple of other credentials which I am not at liberty to put in a public blog post.
Second, the notion of “better” without objective data and specific criteria against which the data were measured is useless. My grandmother’s apple pie was better than your grandmother’s. Nice assertion. Just meaningless. (By the way, I hated my grandmother’s pies, and I avoid pies to this day. Subjective experiences are important, but they are anchored in preferences, prejudice, and personal interpretation of a thing, an experience, or a pie.)
Third, Google search functions today mostly as it has since 2006 when the great drift downwards began as the ad revenues began to soar. You access these “controls” at https://www.google.com/advanced_search. If the idea of using playtime Boolean troubles your modern mind, you can venture into the world of Google dorks, for-fee search services, and zippier dashboards that make information findable with a click. Believe me, consumers don’t want to use these three options because there are more sophisticated ones available for the professional who digs for facts and needs to verify them.
Fourth, fuzzy AI interfaces provide the Google with three extras those looking for groceries don’t think about. These are:
- Data useful for personalizing outputs so that higher cost advertising in different forms can be sold to Google’s “real” customers: Partners, agencies, data brokers, etc.
- A mechanism for weaponizing information to suit Google’s purposes. Do you want a snapshot of Sundar Pichai playing cricket. Check out Google Images and let me know if the presented content has been “shaped.” (Tip: It is and will be as long as Google aims to be the Big Dog in the world of information provision.)
- Google search is a marketing confection. It never has and never could index the world’s information. It has never been concerned with precision and recall because its editorial policies are implicit, variable, and inconsistent. Search has been for decades focused on generating revenue, sidestepping regulatory oversight, and presented in kindergarten colors when it is actually what I described in my book Google Version 2.0: The Calculating Predator.
Okay, I am bored with repeating what l have said ad nauseam for many years. This NYT article is an example of the success of the Google digital mind control. The search is BETTER. Great.
It is not much different from the search service called Backrub before Sergey and Larry got money and saw an opportunity to surf on GoTo.com’s Overture concept. Oh, you don’t know about that? Right. Google search is better.
Stephen E Arnold, May 12, 2026
Comments
Got something to say?
-
- Subscribe to Beyond Search
Feature archive
News archive
-
