The story broke Monday evening: in leaked 2021 footage released by the journalist / activist James O’Keefe (his title shifts depending on where one falls along the political spectrum) IBM CEO Arvind Krishna outlined a bonus structure that at least in part rewarded hiring discrimination in favor of women and specific, preferred races. Arvind further plainly stated “asians” did not count as “diverse,” implying — to my ear — hiring too many white or asian men was a punishable offense at the company. This was made more clear by Paul Cormier, chairman of IBM subsidiary Red Hat, who was also included in the shocking footage. In a proud bit of chest pounding, Paul revealed he actually fired executives for not meeting their race quotas. I’ve reached out to IBM for comment on what appears to be a hiring policy in violation of the Civil Rights Act, but have yet to hear back. 

After the news broke, America First Legal, a non-profit run by former Trump administration official Stephen Miller, filed a federal civil rights complaint against IBM for “racially discriminating against white and Asian Americans and promising to fire, demote, or deny bonuses to corporate executives who fail to meet their illegal race and sex-based hiring quotas.” But from here, legally speaking, it’s not really clear what happens. No former or prospective employee impacted by the policy has come forward with a charge against the company. In fact, we’re still not entirely certain these policies exist. Yes, IBM’s CEO said they exist, but America’s race discourse over the last few years has been absolutely wild, and I guess it’s at least possible Arvind was lying to appease the small handful of deranged employees at his company who worship at the altar of virulent, fantastical race obsession. At this point, his lawyers have likely told him “please stop talking,” and also “literally wtf,” so we’ll just have to wait for the next move. 

I do think, however, a next move is coming.

IBM only received a slight bit of mainstream support. It came from Bloomberg, which predictably framed the company’s overtly racist hiring practices as a “diversity effort” targeted by “James O’Keefe.” But the actual piece pretty much reported on the story accurately, which, to me, looks sort of like exhaustion. Defense of racist DEI policy was probably always a little bit morally taxing, but these days the media has also lost our broader culture on the issue, which means defense of such abhorrent practices is always met with fury. 

Now, almost every day on X, some former fence sitter is denouncing DEI. Serious legal challenges to the present order are inevitable, and the Supreme Court’s ruling over overt racism in higher education seems to have paved the way for massive blows to the corporate DEI kraken. My sense is every executive who has ever engaged in this stuff will be legally liable over the years to come, regardless of their “nice guy” motives, and they’ll no longer be receiving glowing media puff pieces for their trouble.

Good rule of thumb for young founders moving forward: simply don’t be racist.

NOTE: If you’ve been impacted by IBM’s hiring quotas, or any racist policy, at this or any company in the industry, please hit me up (solana@piratewires.com) — I’d love to talk.

-

THE FIFTH ESTATE

NOTABLE INDUSTRY TRENDS

California’s tech-induced (???) fiscal apocalypse, incoming. California’s budget deficit has doubled from 2022, ballooning to an unimaginable $68 billion. In a report released Thursday, the Legislative Analyst’s Office (LAO) cited the following culprits: economic downturn (is that real? How are we measuring this?), higher borrowing costs (lower now than at any point throughout the 70s and 80s, and roughly comparable to what we saw throughout the 90s), lower home sales (what, you don’t want to pay $5 million for a fixer-upper fully equipped with a plucky meth addict in your garage?), and a drastically reduced number of Silicon Valley IPOs (ah yes, I see it’s your fault still). The LAO advised Newsom to declare a fiscal emergency, which would allow him to tap into California’s rainy day funds, though the funds — even if completely tapped — would only cover about 30% of the shortfall. (Politico)

There are two ways this all breaks down: 1) massive service cuts, or 2) tax hikes targeting the state’s productive class, and especially the technology industry, which the state just noted isn’t pumping out golden eggs like it was a couple years ago. The first strategy is rational, but painful, and politically impossible in the Democrat-controlled state of California. The second strategy is irrational, will further accelerate the California doom loop, and is therefore most likely.

Have you had enough of the AI doomsday influencer grift? I have not had enough of the AI doomsday influencer grift. Last week, Gary Marcus, the shameless “pick me” professor nobody asked for comment, who has nonetheless clawed his way into the AI safety discussion, went to war with equally tedious “AI safety” fraud Timnit Gebru (who famously quit Google, then told the world that she was fired by the company). This week, Gary picked a fight with Kara Swisher, who also sucks (though considerably less than the other two, I reluctantly admit). And after the dust cleared, Professor Please Look at Me weirdly posted a list of people he’s tried to start fights with this week on Twitter, thus revealing perfectly the psychology of the typical AI safetyist, which is above all things obsessed with status.

FLASHBACK —

Seven months ago, I deconstructed the first major AI hearing for Pirate Wires in a great piece (or at least according to my mom) called Regulate Me, Daddy. Sam Altman was the star of the show, and from this performance we did glean some important insights into the regulatory battle for AI. But Gary was also there, and he made an amazing spectacle of himself. At the time, I noted the man was clearly overjoyed to be included, an important quality among the “AI safety” people — pageantry, performance, attention — and today I noticed he actually included his few, embarrassing moments before Congress in his Twitter bio.

Lmao bless.

We should expect the AI conversation to carry on in measure roughly this idiotic for the foreseeable future.

INDUSTRY LINKS

BROAD TECH:

• Two weeks after Forbes doxxed Beff Jezos, e/acc made its grand debut in the pages of the New York Times, via Kevin Roose. A reasonably fair portrait from the man last seen having “trouble sleeping” after priming a chatbot to say spooky things, which it then did, which he then wrote about in a 10,000-word AI hit piece. (NYT)

• Bryan Johnson, the Immortal Man, was exonerated by courts regarding a scandal his ex-wife’s lawyers manufactured, which the media naturally and happily amplified. He posted a full account of the ordeal here.

• Elon reinstated Alex Jones’ X account following a poll on the platform in which 70.1% of respondents voted in favor of the gigachad conspiracy theorist’s digital resurrection. This is a stark reversal from Elon’s 2022 stance: “I have no mercy for anyone who would use the deaths of children for gain, politics or fame” (Bloomberg). Civil rights attorney Sherrilyn Ifill announced she’s leaving X in response, somehow still not understanding the concept of ‘free speech on X’ following a year, now, of Elon’s rule. (The Hill)

• Starlink crushing it: global internet traffic on the network nearly tripled in 2023. In Brazil, usage is up 17x from 2022. (Business Insider)

• Possibly Based Commerce Secretary Gina Raimondo says the US will take the “strongest possible” action to protect national security, saying their department has launched a probe following this Fall’s news of breakthroughs in Huawei’s chip technology. (Bloomberg)

• Beeper Mini, an app that allows Android users’ messages to show up as blue bubbles on iPhones, experienced downtime over the weekend as Apple cut off access to Beeper. By Monday, the app was functioning normally again, perhaps because Apple’s lawyers remembered that under the 1998 DMCA legislation reverse engineering for the purpose of interoperability is protected under US law. (TechCrunch)

BIG EPISODE THIS WEEK: Balaji Srinivasan and Beff Jezos joined me on the Pirate Wires pod to discuss Beff’s doxxing by Forbes, tech’s relationship to the media, artificial intelligence, e/acc, decels, and more.

Human resources:

• In an “initiative for students who because of antisemitism fear for their safety on campus and need to seek refuge outside traditional establishments of higher education,” Palantir set aside 180 positions. Loving Karp’s ascendant IDGAF trajectory, but this is not the tea for me. It doesn’t matter that the “yay Hamas” people are evil. It doesn’t matter that the college presidents are all insane. Racial quotas are still, for any reason, a deeply messed up concept. (@PalantirTech)

• 38% of American workers freelanced in 2023, according to a study from Upwork. For Gen Z, that number was 52%. (Bloomberg)

• Google hired its first-ever director of product inclusion, a DEI position “meant to help ensure that people can use what Google offers regardless of where they live, what they earn, how they look or what their bodies can do.” Unclear who previously lacked access to Google, an internet-based suite of products, but go off I guess. (WSJ)

• Tribe Capital’s Arjun Sethi passed the CEO role to Boris Revsin, and assumed the chairman and CIO roles. (Bloomberg)

• GM CEO Mary Barra is calling staff back into the office for 3 days out of the week, ending the company’s COVID-era “Work Appropriately” unlimited WFH policy. (Bloomberg)

• Microsoft is cozying up to unions, launching a new “AI alliance” with the AFL-CIO that will foster information-sharing and give labor leaders and workers education on AI. The alliance also includes terms that would make it easier for unions to organize at Microsoft. (Bloomberg)

• SF-based software analytics company New Relic has listed half of its four floors of office space in downtown SF for sublease. (SF Chronicle)

• Andreessen Horowitz has opened an office in Miami. (@KTmBoyle)

• A Bloomberg headline: “Office Holiday Party Booze Is Being Replaced With Guac and Pickleball.” The piece goes on to say corporate leadership is “opting for year-end celebrations cut totally free from seasonal connotations” (Bloomberg). It truly does seem everything about this trend was designed in a lab to personally annoy me. Pirate Wires just wrapped our Christmas celebration, for what it’s worth, and we continued on in the age-old American tradition of getting drunk with our colleagues and inappropriately sharing details about our personal lives. There was no PICKLEBALL. Ffs. Trash generation.

$$$:

• Cathie Wood’s Ark ETF is loading up on Microsoft and Meta. Her fund has outperformed the Nasdaq 100 so far this year. (Bloomberg)

• France’s Mistral AI has officially closed their $415 million fundraise, which values the company at $2 billion. Thanks to what I can only assume was a masterstroke lobbying effort on the part of France and some VC firms, Mistral is exempt from large swaths of Europe’s latest AI regulation on account of the company is still in the “research and development phase.” (Bloomberg)

• Nvidia has invested in “more than two dozen” AI companies this year, with Dealroom estimating the chipmaker participated in 35 deals in 2023, up 6x from 2022. (FT)

• A Sam Altman-backed crypto startup called Meanwhile is raising a $100m Bitcoin fund that will return 5% in BTC yield to investors, generated from lending the BTC to borrowers. (The Block)

• Harvey, the one-year-old AI-for-legal startup, is in talks to raise $70 to 80 million in a round that would value the company at $700 million post-money. PLEASE BUILD FASTER my lawyers want me poor. (The Information)

Litigation and regulation:

• Monday, an SF jury ruled against Google in a suit brought by Epic Games, saying Google engaged in anticompetitive conduct, harmed Epic, and unlawfully prevented Epic from bypassing Google’s 30% platform fee. After the verdict was read, Epic CEO Tim Sweeney broke into a wide grin and slapped his lawyers on the back. (Bloomberg) (Fortune)

• Tesla responded to a California DMV lawsuit alleging false advertising related to its cars’ Autopilot and Full Self Driving features last Friday. They believe the case has no legal standing because, in effect, the DMV already implicitly approved the use of these names after not caring about them for the past seven years. If Tesla loses the suit, the company’s CA manufacturer’s license could be revoked, and the company could be required to refund customers who’ve paid for the Full Self Driving feature. (LA Times)

• Nikola founder Trevor Milton should spend about a decade in prison, prosecutors argue, saying the sentence would deter others from making false claims to investors. Milton will be sentenced on December 18. (Bloomberg)

• Friends, I am clutching my pearls. Tradeshift’s former CEO responded to totally wild sex slave allegations leveled by a former romantic partner who worked as his assistant, saying the claims constitute defamation. My two cents: it does strike me as a little strange this puritanical, damsel in distress would happily sign a literal sex slave contract if not interested in roleplaying sex slavery, but who knows really. Kids today. I didn’t even think they were having sex! Who can keep up? (SF Standard)

AI:

• Elon’s xAI released its buggy LLM Grok to all X premium subscribers. One user shared a screenshot in which Grok cited an OpenAI policy. Awkward. (Ars Technica)

• Meta released an AI image generator trained on “1.1 billion publicly visible Facebook and Instagram images” called Imagine. (Ars Technica)

• It sounds like Google may have staged elements of its multimodal AI Gemini demo to show capabilities it doesn’t have. Per Engadget, the model’s outputs were sped up, and it was never given commands by a human voice. Instead, they made the demo by “‘using still image frames from the footage, and prompting via text,’ rather than having Gemini respond to — or even predict — a drawing or change of objects on the table in real time.” This is on top of the fact that “Google essentially fudged the benchmark comparisons via the fine print... [they] created a new non standard methodology, then showed that next to OpenAI’s model using the standard method,” according to Founders Fund’s Joey Krug.

• The US Government Accountability Office (GAO) found more than 1,200 potential use cases for AI across all federal agencies. Somehow, and absolutely, this will mean we need more government jobs. (The Hill)

“Friends,” “allies”:

• The EU reached a deal with their AI Act following a “nearly 24 straight hour” negotiation in Brussels (Bloomberg). Stories surrounding the landmark regulatory framework for technology nobody on the continent has built were truly mortifying for this once proud people: they worked so long and hard the self-service coffee machine broke, some said. Finally, hall monitors crowded around an iPad for a Situation Room-style photograph. “Everyone say sleep deprived,” one woman chuckled.

• Emmanuel Macron warns the Act risks stifling innovation on the continent: “We can decide to regulate much faster and much stronger than our major competitors. But we will regulate things that we will no longer produce or invent. This is never a good idea.” (FT)

• THEORY: what if this is all just some elaborate cope? The Europeans were never going to build something — they never do! — and now they have a comfortable reason. It was this idiot in Brussels, you see.

• In any case, decline abroad continues.

• The United Kingdom, which has recently taken to wielding its special trade relationship with the United States against the interests of our country, has decided it’s also mad about the OpenAI-Microsoft deal — two companies it has nothing the fuck to do with. Friday, UK antitrust legislators launched an inquiry into the relationship. (TechCrunch)

• These people are all really annoying! Trade embargo!

Crypto:

• After a TV spot in which she said that crypto is a “new threat” that “we can’t allow to continue,” Elizabeth Warren announced five more senators will co-sponsor her “Digital Asset Anti-Money Laundering Act,” which she reintroduced in July. Crypto advocacy and research non-profit Coin Center calls the bill “the most direct attack on the personal freedom and privacy of cryptocurrency users and developers we’ve yet seen [...] To the extent cryptocurrencies could even continue to exist in a world where this bill becomes law, Americans’ ability to use them would be limited to a fully permissioned and surveilled environment.” (Coin Center)

• Warren’s recent bullshit is just the latest in a long line of crypto assassination attempts by the Biden Administration. Catch up with Nic Carter’s Operation Chokepoint 2.0 piece for Pirate Wires, and his follow-up Did The Government Start A Global Financial Crisis In An Attempt To Destroy Crypto?

• FLASHBACK, and apropos of nothing in particular: Warren lied for decades about her Native American ancestry (this has nothing to do with the crypto stuff I still just find it extremely funny). 

• Fidelity has listed its BTC spot ETF on the Depository Trust & Clearing Corporation with the ticker $FBTC in preparation for an SEC approval window expected between January 5 and January 10 of next year. (Benzinga)

• Montenegro has extended the detention of Do Kwon, co-founder of Terraform Labs, at the request of the US and South Korea. Both countries are seeking charges against him related to the 2022 collapse of TerraUSD. (Bloomberg)

• SBF’s lawyer says the defendant may be “the worst person I’ve ever seen do a cross examination.” (Bloomberg)

• Crypto hacks have netted around $1.7 billion this year, down roughly 50% from last year. (Bloomberg)

Trade war:

• A Chinese invisibility cloak hoax video is once again making the rounds on social media — this time on the country’s spy app TikTok. Portions of this same video made the rounds at least one time previously, in 2017, after Western media picked it up from a viral Weibo post. For all its faults, the website Snopes pretty effectively debunked the most sensational parts of the video in 2017, showing that “the actual footage contains several clues that this invisibility cloak is nothing more than the result of video editing.” Now, much of the older footage is combined with what seems to be one new clip of a Chinese guy unveiling what appears to be a screen that obscures the lower half of his body at some kind of ‘American Idol’ type reality competition, and the TikTok children are panicking. 

• TikTok is the first non-gaming app to generate over $10 billion in consumer spending revenue, according to Data.ai. Tinder and YouTube are closest behind TikTok, having generated nearly $8 billion and over $7 billion respectively. (Data.ai)

• Voter support for a TikTok ban has dropped broadly from Q1 of this year. Just 18% of U.S. teens ages 13 to 17 support a ban, while 49% of American seniors are in favor of one. (The Hill) (Pew Research)

• Guys we really gotta ban this app