The Paris AI Action Summit wrapped up last week with the signing of a joint declaration by 61 countries (excluding the US and the UK) that said nothing very eloquently:
[The Declaration is a] focus on “ensuring AI is open, inclusive, transparent, ethical, safe, secure and trustworthy.” It also calls for greater collaboration when it comes to AI governance, fostering a “global dialogue.”
More concrete and direct was Vice President Vance’s speech pillorying the EU’s regulatory regime—something also espoused by other EU countries like France. The vice president’s message isn’t far off base. It is no coincidence that the EU isn’t home to the Googles and Microsofts of the world, while the US, with a much laxer regulatory regime, is.
However, the EU’s regulatory state may retreat in the future. Defense spending is up.“VC funding in European Deep Tech Defence, Security and Resilience is at an all time high, reaching $5.2B in 2024.” (You can download the full report at the Not Those Shades of Gray Substack). These numbers will probably continue to grow, given the increasingly volatile world. Complex regulations preventing growth and innovation will not survive first contact.
In other security news
US private equity-backed data centres fuel growth of TikTok’s Chinese owner.
“US private equity groups have invested billions of dollars in data centres serving TikTok owner ByteDance, in a dealmaking frenzy now threatened by a US crackdown on Chinese companies’ access to the best chips.”
China tightens grip on tech, minerals and engineers as trade war spirals
“Chinese authorities in recent months have made it more difficult for some engineers and equipment to leave the country, proposed new export controls to retain key battery technologies, and moved to restrict technologies for processing critical minerals, according to multiple industry figures and ministry notices.”
China’s Salt Typhoon hackers continue to breach telecom firms despite US sanctions
“In a report shared with TechCrunch, threat intelligence firm Recorded Future said it had observed Salt Typhoon — which the company tracks as “RedMike” — breaching five telecommunications firms between December 2024 and January 2025.”
Paywalls vs. “zero-click future”
AI means the end of internet search as we’ve known it
The biggest change to the way search engines have delivered information to us since the 1990s is happening right now. No more keyword searching. No more sorting through links to click. Instead, we’re entering an era of conversational search. Which means instead of keywords, you use real questions, expressed in natural language. And instead of links, you’ll increasingly be met with answers, written by generative AI and based on live information from all across the internet, delivered the same way.
The shift to “conversational searching” will fundamentally change how we interact with the internet. I’m usually hesitant to engage in hyperbolic future predicting, but there are a few reasons why I’ll engage with it now.
1. “Zero Click” future.
The article linked above describes how publishers of web pages are worried. Publishers receive revenue by driving people to their websites. This revenue can manifest in many ways: ad revenue, data collected by trackers, and actual sales of goods and services. If no one goes to the website, many of those revenue streams are gone. Providers of information–journalists, academics, and substack writers—for instance—will be the hardest hit. While most of my user engagement comes from Substack itself and my occasional LinkedIn posts, I am receiving increased Perplixity traffic. I’ve also yet to convert a Perplexity viewer into a subscriber. Perhaps another method will emerge that will drive users to websites, or maybe in the future, we will all just be writing for the LLMs.
2. LLM echo chambers
Google search has undoubtedly changed over the years. Websites are driven to the top for various reasons, sometimes through popularity and merit, other times through cleverly engineered Search Engine Optimization. Yet indexing is the digital equivalent of a card catalog. LLMs that provide answers to search queries are forced to take “editorial stances” on how they present answers to controversial topics. Imagine walking into the library looking for a book on the Civil War, and instead of seeing a shelf of historical treaties from all sides of the battle, you are only given a book on the Lost Cause. This isn’t a new problem, but it will become an exasperated one when AI searches are the only game in town.
3. Paywalls, Paywalls, Paywalls.
The internet has thus far avoided turning into a pay-per-view version of itself, but new search methods may change that. “Crawlers,” bots that roam the web, are responsible for indexing websites for Google, training AI models, and research. They account for half of all internet traffic, and because of their role in training AI by crawling and copying massive amounts of data without permission, they are being turned away through various means, sometimes through code and sometimes through paywalls. What does this mean?
Large websites and publishers will defend their content in court or negotiate contracts. And massive tech companies can afford to license large data sets or create powerful crawlers to circumvent restrictions. But small creators, such as visual artists, YouTube educators, or bloggers, may feel they have only two options: hide their content behind logins and paywalls, or take it offline entirely. For real users, this is making it harder to access news articles, see content from their favorite creators, and navigate the web without hitting logins, subscription demands, and captchas each step of the way.
In other news
Dating App Cover-Up: How Tinder, Hinge, and Their Corporate Owner Keep Rape Under Wraps. A thorough report of the lack of safety features on Dating apps, profiling a case where a man convicted of “sexually assaulting eight women, drugging two women, and assaulting one more for a total of 11 women.” remained on the app despite multiple reports of abuse and criminal investigations. The company at the center of the investigation, Match Group, is currently trying to use AI “to detect ‘off-colour’ messages” in an effort to make their platform more appealing to women.
TAKE IT DOWN Act - S.146 was passed by the Senate by unanimous consent. “The act would require covered platforms to remove nonconsensual intimate visual depictions. It aims to combat the nonconsensual sharing of intimate visual depictions, including "deepfakes," by establishing criminal penalties for violators and by creating a process for victims to request the removal of such content from online platforms.”
Space flight launches turn to the sea as coastal land runs out and regulations hinder launches. “I can imagine that some unauthorized projects may become possible simply because they are on the seas and there is no real authority—by contrast to land-based space launches—to supervise those kinds of launches,”
“X to pay Donald Trump $10m to settle lawsuit over Capitol attack”
“Death of OpenAI whistleblower deemed suicide in new autopsy report”