davel

Aside from Alphabet’s own forays into AI, they’re also selling pickaxes to most other AI prospectors.

It’s a story deSouza likes to tell in numbers. In a conversation with this editor, he notes several times that nine out of the top 10 AI labs use Google’s infrastructure. He also says that nearly all generative AI unicorns run on Google Cloud, that 60% of all GenAI startups worldwide have chosen Google as their cloud provider, and that the company has lined up $58 billion in new revenue commitments over the next two years, which represents more than double its current annual run rate.

Asked what percentage of Google Cloud’s revenue comes from AI companies, he offers instead that “AI is resetting the cloud market, and Google Cloud is leading the way, especially with startups.”

The strategy extends beyond simple customer acquisition. Google offers AI startups $350,000 in cloud credits, access to its technical teams, and go-to-market support through its marketplace. Google Cloud also provides what deSouza describes as a “no compromise” AI stack — from chips to models to applications — with an “open ethos” that gives customers choice at every layer.

The approach reflects both opportunity and necessity. In a market where companies can go “from being a startup to being a multibillion-dollar company in a very short period of time,” as deSouza puts it, capturing future unicorns before they mature could prove more valuable than fighting over today’s giants.

“Companies love the fact that they can get access to our AI stack, they can get access to our teams to understand where our technologies are going,” deSouza says during our interview. “They also love that they’re getting access to enterprise-grade Google class infrastructure.”

Google’s infrastructure play got even more ambitious recently, with reporting revealing the company’s behind-the-scenes maneuvering to expand its custom AI chip business. According to The Information, Google has struck deals to place its tensor processing units (TPUs) in other cloud providers’ data centers for the first time, including an agreement with London-based Fluidstack that includes up to $3.2 billion in financial backing for a New York facility.

Competing directly with AI companies while simultaneously providing them infrastructure requires … finesse. Google Cloud provides TPU chips to OpenAI and hosts Anthropic’s Claude model through its Vertex AI platform, even as its own Gemini models compete head-to-head with both. (Google Cloud’s parent company, Alphabet, also owns a 14% stake in Anthropic, per New York Times court documents obtained earlier this year, though when asked directly about Google’s financial relationship with Anthropic, deSouza calls the relationship a “multi-layered partnership,” then quickly redirects me to Google Cloud’s model marketplace, noting that customers can access various foundation models.)

It’s hard to overstate how much different NSPM-7 is from the over 200 executive orders Trump has frantically signed since coming back into office.

NSPM-7 directs a new national strategy to “disrupt” any individual or groups “that foment political violence,” including “before they result in violent political acts.”

In other words, they’re targeting pre-crime, to reference Minority Report.

The Trump administration isn’t only targeting organizations or groups but even individuals and “entities” whom NSPM-7 says can be identified by any of the following “indica” (indicators) of violence:

• anti-Americanism,
• anti-capitalism,
• anti-Christianity,
• support for the overthrow of the United States Government,
• extremism on migration,
• extremism on race,
• extremism on gender
• hostility towards those who hold traditional American views on family,
• hostility towards those who hold traditional American views on religion, and
• hostility towards those who hold traditional American views on morality

It may not work in Safari, and it seems to have been optimized for Chromium browsers.

As an avid (ab)user of phrases-as-lemmata, I found this interesting.

Discovered through YouTube: Linguists just made a breakthrough in defining a 'word. ' No, really

[Swedish] count [Eric von Rosen] used the swastika as a personal good luck charm. When he gifted a plane to the nascent air force of Sweden's newly independent neighbour in 1918 he had had a blue swastika painted on it. This Thulin Typ D was the first aircraft of the Finnish air force and subsequent planes all had his blue swastika symbol too, until 1945.

Supporters of a continued use of the symbol point out that there were no Nazis in 1918 so the air force's use of the swastika has nothing to do with Nazism.

However, while Eric von Rosen had no Nazi associations at the time of his 1918 gift, he did subsequently become a leading figure in Sweden's own national socialist movement in the 1930s. He was also a brother-in-law of senior German Nazi Herman Göring, and, according to Prof Teivainen, a personal friend of Hitler.

So the fascists adopted the swastika by way of a Swedish Count-cum-fascist.

Also: https://en.wikipedia.org/wiki/Air_Force_Academy_(Finland)

Unicode is good. If you’re designing a data structure or protocol that has text fields, they should contain Unicode characters encoded in UTF-8. There’s another question, though: “Which Unicode characters?” The answer is “Not all of them, please exclude some.”

This issue keeps coming up, so Paul Hoffman and I put together an individual-submission draft to the IETF and now (where by “now” I mean “two years later”) it’s been published as RFC 9839. It explains which characters are bad, and why, then offers three plausible less-bad subsets that you might want to use. Herewith a bit of background, but…

Please · If you’re actually working on something new that will have text fields, please read the RFC. It’s only ten pages long, and that’s with all the IETF boilerplate. It’s written specifically for software and networking people.

Source code · I’ve written a little Go-language library to validate incoming text fields against each of the three subsets that 9839 specifies, here. I don’t claim it’s optimal, but it is well-tested.

Details · Here’s a compact summary of the world of problematic Unicode code points and data formats and standards.

Notes:
[1] XML allows C1 controls.
[2] XML and YAML don’t exclude the noncharacters outside the Basic Multilingual Pane.
[3] YAML excludes all the legacy controls except for the mostly-harmless U+0085, another version of \n used in IBM mainframe documents.

On August 7th, US polling giant Gallup published the remarkable results of a survey of Ukrainians. Public support for Kiev “fighting until victory” has plummeted to a record low “across all segments” of the population, “regardless of region or demographic group.” In a “nearly complete reversal from public opinion in 2022,” 69% of citizens “favor a negotiated end to the war as soon as possible.” Just 24% wish to keep fighting. However, vanishingly few believe the proxy war will end anytime soon.

The reasons for Ukrainian pessimism on this point are unstated, but an obvious explanation is the intransigence of President Volodymyr Zelensky, encouraged by his overseas backers - Britain in particular. London’s reverie of breaking up Russia into readily-exploitable chunks dates back centuries, and became turbocharged in the wake of the February 2014 Maidan coup. In July that year, a precise blueprint for the current proxy conflict was published by the Institute for Statecraft, a NATO/MI6 cutout founded by veteran British military intelligence apparatchik Chris Donnelly.

In response to the Donbass civil war, Statecraft advocated targeting Moscow with a variety of “anti-subversive measures”. This included “economic boycott, breach of diplomatic relations,” as well as “propaganda and counter-propaganda, pressure on neutrals.” The objective was to produce “armed conflict of the old-fashioned sort” with Russia, which “Britain and the West could win.” While we are now witnessing in real-time the brutal unravelling of Donnelly’s monstrous plot, Anglo-American designs of using Ukraine as a beachhead for all-out war with Moscow date back far further.

In August 1957, the CIA secretly drew up elaborate plans for an invasion of Ukraine by US special forces. It was hoped neighbourhood anti-Communist agitators would be mobilized as footsoldiers to assist in the effort. A detailed 200-page report, Resistance Factors and Special Forces Areas, set out demographic, economic, geographical, historical and political factors throughout the then-Soviet Socialist Republic that could facilitate, or impede, Washington’s quest to ignite local insurrection, and in turn the USSR’s ultimate collapse.

[…]

‘Strongly Anti-Nationalist’

The CIA’s invasion plan never formally came to pass. Yet, areas of Ukraine forecast by the Agency to be most welcoming of US special forces were precisely where support for the Maidan coup was highest. Moreover, in a largely unknown chapter of the Maidan saga, fascist Right Sector militants were bussed en masse to Crimea prior to Moscow’s seizure of the peninsula. Had they succeeded in overrunning the territory, Right Sector would’ve fulfilled the CIA’s objective, as outlined in Resistance Factors and Special Forces Areas.

A civilian defence barricade constructed to prevent Right Sector entering Crimea, February 2014

Given what transpired elsewhere in Ukraine following February 2014, other sections of the CIA report take on a distinctly eerie character. For instance, despite its strategic position facing the Black Sea, the Agency warned against attempting to foment anti-Soviet rebellion in Odessa. The agency noted the city is “the most cosmopolitan area in Ukraine, with a heterogeneous population including significant numbers of Greeks, Moldovans and Bulgarians, as well as Russians and Jews.” As such:

“Odessa…has developed a less nationalistic character. Historically, it has been considered more Russian than Ukrainian territory. There was little evidence of nationalist or anti-Russian sentiment here during the Second World War, and the city…was in fact controlled by a strongly anti-nationalist local administration [during the conflict].”

Odessa became a key battleground between pro- and anti-Maidan elements, from the moment the protests erupted in November 2013. By March the next year, Russophone Ukrainians had occupied the city’s historic Kulykove Pole Square, and were calling for a referendum on the establishment of an “Odessa Autonomous Republic”. Tensions came to a head on May 2nd, when fascist football ultras - who subsequently formed Azov Battalion - stormed Odessa and forced dozens of anti-Maidan activists into Trade Unions House, before setting it ablaze.

In all, 42 people were killed and hundreds injured, while Odessa’s anti-Maidan movement was comprehensively neutralised. In March this year, the European Court of Human Rights issued a damning ruling against Kiev over the massacre. It concluded local police and fire services “deliberately” failed to respond appropriately to the inferno, and authorities insulated culpable officials and perpetrators from prosecution despite possessing incontrovertible evidence. Lethal “negligence” by officials on the day, and ever after, was found to go far “beyond an error of judgment or carelessness.”

The ECHR was apparently unwilling to consider the incineration of anti-Maidan activists was an intentional and premeditated act of mass murder, conceived and directed by Kiev’s US-installed fascist government. However, the findings of a Ukrainian parliamentary commission point ineluctably towards this conclusion. Whether, in turn, the Odessa massacre was intended to trigger Russian intervention in Ukraine, thus precipitating “armed conflict of the old-fashioned sort” with Moscow that “Britain and the West could win” is a matter of speculation - although the Institute for Statecraft was present in the country at the time.