Save-Point

Wozniak, Musk, and more than 1000 other business leaders signed a letter seeking guardrails and a pause on training AI models as the technology grows more powerful. The letter argues powerful AI models like OpenAI's GPT-4 "should only be developed only once we are confident that their effects will be positive and their risks will be manageable."

James Grimmelmann, professor of digital and information law at Cornell University Wrote:It's a very good idea to slow down development of new models because if AI ends up being good for us, then there's no harm waiting months or years, we'll get to the end point anyway. If it's harmful, then we just bought ourselves extra time to strategize the best ways to respond and understand how to combat it.

It is true that ChatGPT was a blockbuster achievement, and its capabilities have been pushed to the limits. But it has now been used to pass medical licensing exams, gave instructions how to make explosives, and created an alter ego for itself. AI products including ChatGPT have already been shown to misrepresent information and make mistakes, something that Steve Wozniak had spoken about in public.  And even OpenAI CEO Sam Altman admitted that his company's model could spit out racist, sexist and completely biased answers.



 This kinda reminds me of why DUNE has no computers (not even a calculator if you never noticed):
... from the glossary at the end of the 500+ page novel ...
Predating THE MATRIX as it was written in 1967, DUNE has a backstory of AI taking control over mankind and a war that had to be waged for the freedom of humanity against the machines.

Quote:Efforts from Sen. Josh Hawley (R-Mo.) to ban Chinese-owned TikTok on U.S. soil suffered a setback on March 29 when fellow Republican Sen. Rand Paul (Ky.) opposed the move, arguing the ban would limit free speech.

On the Senate floor, Paul objected to Hawley’s request for unanimous consent to fast-track a ban on the viral app in the United States over national security concerns, telling his fellow colleagues he sees the ban as a First Amendment issue.

“Before banning TikTok, these censors might want to discover that China’s government already bans TikTok. Hmmm … do we really want to emulate China’s speech bans?” he said.

While TikTok does not operate in China, Douyin, the app’s heavily-censored sister platform run by TikTok’s Chinese parent company ByteDance, is allowed to operate in the country.

“Have faith that our Americans are smart enough to hear bad ideas and reject those ideas, have faith that our desire for freedom is strong enough to survive a few dance videos.”
...
Responding to Paul, Hawley said that the First Amendment doesn’t contain the “right to espionage.”

“I’m astounded to learn that Americans have the right to be spied upon,” he said in a speech.

Who called for a spy?  

Quote:“Apple has released security updates to address vulnerabilities in multiple products. An attacker could exploit some of these vulnerabilities to take control of an affected device,” the notice, dated March 28, reads. “CISA encourages users and administrators to review the following advisories and apply the necessary updates.”

Security updates were included in the iOS 15.7.4 and iPadOS 15.7.4, iOS 16.4 and iPadOS 16.4, Safari 16.4, Studio Display Firmware Update 16.4, watchOS 9.4, tvOS 16.4, macOS Big Sur 11.7.5, macOS Monterey 12.6.4, and macOS Ventura 13.3 upgrades on Monday.

Apple Inc. on Monday confirmed it released its iOS 16.4 update to users, including a number of new security updates and features.

iOS 16.4 runs on all iPhones starting from the iPhone 8 onwards, according to Apple, which includes bug fixes and new features. Apple also released iOS 15.7.4 for iPhones that are older who have older devices Monday.

The iOS 16.4 upgrade also fixes two flaws in the iPhone operating systems, tracked as CVE-2023-27969 and CVE-2023-27933 that could allow a malicious actor to execute code without the user knowing. Meanwhile, another flaw, CVE-2023-28178, could allow an attacker to bypass the user’s Privacy preferences, says Apple via its support page.

Two vulnerabilities in WebKit, which powers the Safari browser, were also fixed in the latest update, the release notes says. Overall, more than 30 security issues were fixed in the update.

Quote:Bill Gates recently praised the evolution of artificial intelligence, his relationship with OpenAI, and gave a short warning on the situation being portrayed differently by other subject experts, including Elon Musk.

The Microsoft co-founder started out his March 21 GatesNotes post on AI in a hopeful tone: “The development of AI is as fundamental as the creation of the microprocessor, the personal computer, the Internet, and the mobile phone. It will change the way people work, learn, travel, get health care, and communicate with each other. Entire industries will reorient around it. Businesses will distinguish themselves by how well they use it.”

Gates said that AI can help with several progressive agendas, including climate change and economic inequities, but that the technology is “disruptive,” and will “make people uneasy.”

“AIs also make factual mistakes and experience hallucinations.” AI hallucinations are confident responses by an AI that are not grounded in its training data. Frequent hallucinations are considered to be a major issue with large language models like ChatGPT.

“In addition, advances in AI will enable the creation of a personal agent. Think of it as a digital personal assistant: It will see your latest emails, know about the meetings you attend, read what you read, and read the things you don’t want to bother with,” Gates said.
...
Regarding the education sector, Gates said: “It will know your interests and your learning style so it can tailor content that will keep you engaged. It will measure your understanding, notice when you’re losing interest, and understand what kind of motivation you respond to. It will give immediate feedback.”

Gates starts this section with the fact that AI does not understand “context for a human’s request,” leading to “strange results.” For example, “when you ask for advice about a trip you want to take, it may suggest hotels that don’t exist.”

Although such technical issues will get resolved, there are some problems that pose a greater danger.

“For example, there’s the threat posed by humans armed with AI. Like most inventions, artificial intelligence can be used for good purposes or malign ones.”

He then added, “Then there’s the possibility that AIs will run out of control. Could a machine decide that humans are a threat, conclude that its interests are different from ours, or simply stop caring about us?”

 That makes it hard to properly discern if the major cyber threat humanity is facing right now is OpenAI aka ChatGPT or Bill Gates himself.

Quote:JPMorgan Chase & Co. is planning to test new technology that would let consumers pay with their palms or faces at certain US merchants.

The bank, home to one of the world’s biggest payment-processing businesses, plans to roll out the service to its broader base of US merchant clients if the pilot program goes well, according to a statement Thursday. The pilot may include a Formula 1 race in Miami as well as some brick-and-mortar stores.
...
JPMorgan is seizing on the rising popularity of biometrics technology, which uses unique body measurements to authenticate a person’s identity. The technology is expected to account for roughly $5.8 trillion in transactions and 3 billion users by 2026, JPMorgan said, citing Goode Intelligence.

Here’s how it works: Customers enroll their palm or face through an in-store process. Then, at checkout, they scan their biometric to complete the transaction and get a receipt.

The new offering is from JPMorgan’s sprawling payments business, which competes with the likes of Fidelity National Information Services Inc. and Fiserv Inc. to offer payment-processing technology to e-commerce sites and retailers around the world. The technology isn’t just for JPMorgan cardholders. Consumers will be able to link many different forms of payment to JPMorgan’s new system.

 To me it sounds pretty much like a good candidate for the mark of the beast. I doubt most people would ever oppose it for they'll claim it's a highly convenient payment method for everybody, including people and businesses.

Quote:A woman in northern France is to be put on trial on charges of insulting President Emmanuel Macron after describing him as 'filth' in a Facebook post, a prosecutor said on Wednesday. The woman risks a fine of 12,000 euros but not prison if convicted at the trial due to be held in June.

She was arrested on Friday and held in custody for questioning after the state's local administrative office filed a complaint over her Facebook post, the prosecutor in the northern town of Saint Omer, Mehdi Benbouzid, told AFP.

The complaint focused on a post on her Facebook page made on March 21, the day before Macron gave a lunchtime interview to TF1 television to defend his controversial pension reforms that have sparked nationwide protests.

"This piece of filth is going to address you at 1:00 pm... it's always on television that we see this filth," she wrote.

The woman, in her 50s, had been a supporter of the 2018-2019 "Yellow Vest" protests that shook Macron during his first mandate.

She stands accused of "insulting the president of the republic" and will stand trial on June 20 in Saint Omer, the prosecutor said.

"They want to make an example of me," the woman told La Voix du Nord regional newspaper which first reported the accusations.

Quote:A man who posted memes about the Hillary Clinton campaign during the 2016 presidential election was found guilty of one charge of conspiring against the right to vote on Friday.

“Douglass Mackey, also known as ‘Ricky Vaughn,’ was convicted today by a federal jury in Brooklyn of the charge of Conspiracy Against Rights stemming from his scheme to deprive individuals of their constitutional right to vote,” reads a statement released by the Eastern District of New York on Friday.

Members of the jury reached the unanimous verdict after a little less than a week of deliberation starting Monday afternoon after a weeklong period of arguments.

Mackey’s online persona, “Ricky Vaughn”—a reference to the sports comedy film “Major League”—had an online following of over 50,000 during the 2016 presidential election that frequently amplified his pro-Republican voice during the 2016 elections.

Mackey was on trial for a charge of allegedly conspiring against people’s right to vote with memes—or generally satirical online images—including one depicting an African American woman standing in front of an “African Americans for Hillary” sign. The meme instructed people to vote by text during the 2016 presidential election.
...
Mackey’s case is a historical first, in that alleged falsehoods are being argued as a form of election interference, Eugene Volokh, a professor at UCLA School of Law specializing in First Amendment Law.
...
Mackey will be sentenced on Aug. 16, 2023, and faces a maximum of 10 years in prison.
...
Authorities arrested Mackey in January 2021 on that allegation, and a grand jury indicted him within two weeks of his arrest.

According to the prosecutors’ evidence presented in court, Mackey’s Twitter posts included one that suggested limiting “black turnout,” a statement that prosecutors said supports the charge that Mackey intended to conspire to interfere with the election. On the witness stand, Mackey described it as an “exaggeration.”

But according to Mackey’s defense attorney, Andrew Frisch, that meme and Mackey’s Twitter posts were obvious jokes and cannot be taken as a serious attempt at conspiring against the opposition to Mackey’s preferred political party.
...
Although a material injury—in this case, people’s right to vote being taken away—is not required to establish the conspiracy charge, the government said during the trial that people’s votes were “vaporized” by Mackey’s “digital flyers.”
...
Mackey’s defense lawyer said the 4,900 texts were sent after Mackey’s Twitter account was suspended and following media reports of Mackey’s memes. Frisch said that the 4,900 texts were sent because of media reports on Mackey’s memes rather than in response to Mackey’s memes.

During the trial, testimony presented by the prosecutors showed that members of the Clinton campaign reported Mackey’s memes to upper-level campaign managers during the 2016 election cycle, but the managers dismissed their concerns. Prosecutors, when confronted by Frisch about this evidence, did not dispute the factual validity of this evidence but noted the Clinton campaign personnel’s evidence should not be considered as evidence for or against the conspiracy charge.

Frisch, on the other hand, said during the trial that this evidence shows that the government withheld exculpatory evidence, citing Brady v. Maryland, a 1963 case in which the Supreme Court held that prosecutors must make available exculpatory evidence to defense counsel. Frisch said that the government’s withholding of evidence irreparably damaged his opening statements at trial and requested that the court declare a mistrial.

Quote:Voice cloning is becoming increasingly common and with remarkable accuracy through the use of AI technology. However, as with any technological advancement, there will always be those who seek to use it for malicious motives.

In this case, scammers are now using voice cloning to trick you into believing that you are receiving calls from your loved ones or friends when in reality, you are being targeted by a fraudulent scheme.
...
Voice cloning is the creation of an artificial simulation of a person’s voice using artificial intelligence technology. Since a scammer only needs a short amount of recorded speech to clone your voice, they can easily steal your voice and use it for whatever means they wish.

Cybercriminals can do this is by calling you. If you answer a robocall and speak to a scammer, they can take that recording of your voice and use it as a way to mimic you. They can also get your voice through your social media, as many of you take videos of yourselves and post them to Facebook, Twitter, and Instagram for the world to see.

Once they have a voice, they can contact a family member or friend of the person's voice they stole and pretend to be that person. That way, the person on the other line thinks it's someone they know and will more readily agree to give money or other personal information over the phone.

One way to do this is to never answer the phone number of someone you do not know. Spam calls are the perfect way for hackers to get a hold of your voice, and they only need to keep you on the phone for a few minutes.

Be careful what you post online. I know we all love sharing videos of good times with loved ones on our social accounts, however, you should consider making your account or those specific posts private so that only people you're friends with can see them.

Also, consider removing your personalized voicemail from your cell phone and go with a system greeting or default greeting message so that scammers can't steal your voice from your voicemail message and use it for a voice-cloning scam.

Quote:Midjourney, an AI image generator that creates realistic deepfakes, has been scrutinized recently for having a policy showing deference to China's communist government.

The company enforces a rule that users can generate fake images of world leaders from President Biden to Vladimir Putin, but not Chinese President Xi Jinping.

In a year-old message on the chat service Discord, the CEO of Midjourney, Inc. explained why the company has that rule.

"I think we want to minimize drama," Midjourney CEO David Holz wrote last summer. He explained that the company did not immediately ban images of Xi, but it was triggered by abuse from users.

"Political satire in China is pretty not-okay and at some point would endanger people in China from using the service," he added.

"I think people in China using this tech moves a needle in the world in general (in a positive direction)," Holz argued. "i think random people on here doing Chinese political satire does very little to add to anything."

The rule against creating deepfakes of Xi applies to all global users, not just those in China.

Darn policies! Now I can't use that tool to mock him by making him look like Winnie the Pooh.

Quote:Activists and organizations are sounding the alarm that the RESTRICT Act, touted to stop foreign spying via apps like TikTok, will instead endanger basic American freedoms.

A bipartisan group of senators led by Sen. Mark Warner, D-Va., and John Thune, R-S.D., unveiled the RESTRICT Act on March 7. The legislation is meant to crack down on communications technology developed by foreign adversaries, like China and Russia, because of national security risks.

The RESTRICT Act gives the executive branch the power to "[enforce] any mitigation measure to address any risk" regarding a "current, past, or potential future transaction" with what is deemed to be a foreign adversary. It would also apply to taking action "to address any risk arising from any covered transaction by any person, or with respect to any property, subject to the jurisdiction of the United States," including "interfering in, or altering the result or reported result of a Federal election." The penalty for running afoul of this law could be up to "20 years" spent in prison.

Many of the bill's detractors suggest that while it may be touted as a ban on TikTok, it would have far larger implications for civil liberties in the United States.

Sen. J.D. Vance, R-Ohio, wants to ban TikTok but shopped short of supporting the RESTRICT Act, citing comparisons to the Patriot Act.

"I think we should ban TikTok. I'm a little bit more concerned with the RESTRICT Act," he said, according to Business Insider, noting that some people are "very worried that you're creating, effectively, a PATRIOT Act for the digital age."
...
Others online warned about "domestic spying" and the government accessing personal devices.

State Freedom Caucus Network communications director Greg Price condemned the legislation as "absolutely terrifying" in a Twitter thread, warning, "It gives the government the ability to go after anyone they deem as a national security risk at which point they can access everything from their computer to video games to their ring light. This is a Patriot Act for the internet."

Quote:The New York Times lost its verified check mark on Sunday after Twitter CEO Elon Musk announced efforts to crack down on users avoiding payments for "Twitter Blue."

Musk along with the official Twitter Verified account announced on Mar. 23 that the social media company would set a deadline of Apr. 1 for verified users to apply and keep their status. Those who refused to pay the $8 per month subscription for individuals or $1,000 per month for organizations would begin to lose both their blue check mark and verified status.

In a tweet reply, Musk revealed that this could include the New York Times should it refuse to pay for a subscription. Shortly afterward, the publication’s main Twitter account lost its verified status.

"The real tragedy of @NYTimes is that their propaganda isn’t even interesting," Musk tweeted.

He added, "Also, their feed is the Twitter equivalent of diarrhea. It’s unreadable. They would have far more real followers if they only posted their top articles. Same applies to all publications."

Though he did not elaborate on the New York Times losing its verified status, he later called the media outlet "hypocritical" for insisting on people to pay for their subscription while refusing to do the same for Twitter.
...
A New York Times tech article reiterated on Friday that the publication would not pay for its verified badge nor reimburse their journalists for the status.

Quote:The Australian government has banned the social media app Tiktok from all government devices following a security review by Home Affairs Minister Claire O’Neil amid ongoing international concern.

The ban, which was announced on April 4 by Australia’s Attorney General Mark Dreyfus, will prohibit any government-issued devices used by politicians and public servants from installing the social media app.

Dreyfus said in a statement that the ban would come into effect “as soon as practicable.” He said exemptions would be granted on a case-by-case basis.

“After receiving advice from intelligence and security agencies, today I authorised the secretary of the Attorney-General’s Department to issue a mandatory direction under the Protective Security Policy Framework to prohibit the TikTok app on devices issued by Commonwealth departments and agencies,” Dreyfus said.

“The direction will come into effect as soon as practicable.”

However, he noted that possible exemptions will be granted on a “case-by-case basis and with appropriate security mitigations in place.”

The move brings Australia into line with its Five Eyes allies and nine other countries and multinational organisations, including the UK, the United States, India, Canada, New Zealand, Taiwan, the Netherlands, France, Denmark, Norway, Pakistan and the EU.

Concerns around Tik Tok security have risen after the company admitted last year that employees of its parent company, ByteDance, spied on U.S. journalists from Forbes who were investigating the company.

Quote:National Public Radio (NPR) was listed by Twitter as a “US state-affiliated media” on Tuesday, drawing outrage from staffers at the outlet.

Twitter owner Elon Musk weighed in on the designation, writing that it “seems accurate” to call NPR “state-affiliated.” He pointed to Twitter’s Help Center, which noted that “state-affiliated media is defined as outlets where the state exercises control over editorial content through financial resources, direct or indirect political pressures, and/or control over production and distribution.”

And, according to Twitter’s guidelines, “Labels on state-affiliated accounts provide additional context about accounts that are controlled by certain official representatives of governments, state-affiliated media entities, and individuals associated with those entities.”

But staffers with NPR—which critics say exhibits a far-left-wing bias in its reporting—suggested Tuesday the outlet shouldn’t be listed as such. Climate and energy correspondent Jeff Brady dismissed the label by simply writing: “Uh, no.”

Quote:A North Korean government-backed threat actor has been linked to attacks targeting government and military personnel, think tanks, policy makers, academics, and researchers in South Korea and the U.S.
...
Attack chains mounted by ARCHIPELAGO involve the use of phishing emails containing malicious links that, when clicked by the recipients, redirect to fake login pages that are designed to harvest credentials.

These messages purport to be from media outlets and think tanks and seek to entice targets under the pretext of requesting for interviews or additional information about North Korea.

"ARCHIPELAGO invests time and effort to build a rapport with targets, often corresponding with them by email over several days or weeks before finally sending a malicious link or file," TAG said.

The threat actor is also known to employ the browser-in-the-browser (BitB) technique to render rogue login pages inside an actual window to steal credentials.

What's more, the phishing messages have posed as Google account security alerts to activate the infection, with the adversarial collective hosting malware payloads like BabyShark on Google Drive in the form of blank files or ISO optical disc images.

Another notable technique adopted by ARCHIPELAGO is the use of fraudulent Google Chrome extensions to harvest sensitive data, as evidenced in prior campaigns dubbed Stolen Pencil and SharpTongue.

Quote:One of the most popular WordPress plugins, Elementor Pro, used by over eleven million websites, is vulnerable to a high-severity vulnerability that hackers have actively exploited.

More than 12 million sites powered by WordPress have been affected by the vulnerability, which carries a severity rating of 8.8 out of 10.

Elementor Pro is a plugin that allows users to build professional-looking websites without knowing how to code. It provides drag-and-drop functionality like:

• Theme building
  
• A template collection
  
• Custom widget support
  
• WooCommerce support
  

This critical vulnerability occurred in Elementor Pro version 3.11.6. It does, however, allow any authenticated user to update any WordPress setting that has been set on the site.

To accomplish this, an AJAX action within Elementor Pro is used that does not have the proper privilege control in place.

So, it’s strongly recommended that users must update their Elementor Pro plugin to version 3.11.7, released on March 22, 2023, in conjunction with the WooCommerce plugin running on the site.

Quote:The social media platform announced on April 7 that it had updated its abuse and harassment policy to clarify how it defined “targeted harassment.”

There have been increasing complaints from individuals about being attacked and harassed by random Twitter users for their opinions in an attempt to have them driven from the app.

Twitter has responded to complaints by updating the criteria for “targeted harassment” of individuals on its website.

“We believe in free speech and we also believe users have a right to use and enjoy our platform without being subjected to targeted and repeated harassment,” Twitter said.

“We define ‘targeted harassment’ as behavior that is repeated, unreciprocated, and intended to humiliate or degrade an individual(s). This includes targeting people based on gender, race, religion, or sexual orientation.”
...
The relaxation of censorship on Twitter has led to a more relaxed and opinionated user base, but harassment and threats by extremists and left-wing social justice warriors have increased.

The rapper and podcaster, Zuby, tweeted about the various attacks and threats he receives on Twitter.

Some angry users even filed false allegations against certain individuals to Twitter monitors to get them banned or suspended.

The artist said that many of the personal attacks are by people who have no other reason than take glee at ruining his reputation but veil their intent under the guise of virtue signaling

Quote:Hackers have released 16,000 Tasmanian education department documents on the dark web including school children’s personal information, the state government has confirmed.

Science and Technology Minister Madeleine Ogilvie said thousands of financial statements and invoices containing names and addresses of school students and their parents had been released after third-party file transfer service GoAnywhere MFT was hacked.

It is understood the information was released on the dark web by a Russian-linked hacker group called Cl0p.

“At this point in time, I’m advised that there are approximately 16,000 documents that have been released,” Ogilvie told reporters in Hobart on April 7.

“This data has been accessed through a third party file transfer service, and as I have said previously there is no evidence that Tasmanian government IT systems have been breached.

“The information released includes financial invoices and statements, including information relating to student assistance applications. This may include names and addresses and this is an evolving situation.”

The state government has set up a helpline for Tasmanians who are concerned their data has been compromised.

Quote:Google this week announced that users should update their Android smartphones as soon as possible due to multiple vulnerabilities that can lead to “remote code execution.”

In a bulletin issued in April 3, “The most severe of these issues is a critical security vulnerability in the System component that could lead to remote (proximal/adjacent) code execution with no additional execution privileges needed.”

Further, it warned that “user interaction is not needed for exploitation” and that “the severity assessment is based on the effect that exploiting the vulnerability would possibly have on an affected device, assuming the platform and service mitigations are turned off for development purposes or if successfully bypassed.”

The vulnerabilities currently impact Android systems that are using versions 11, 12, 12L, and 13, Google’s bulletin said.

Security experts with antivirus software maker MalwareByes warned that users should “update as soon as they can,” according to a post. “If your Android is on security patch level 2023-04-05 or later, this will address all of these issues. Android partners are notified of all issues at least a month before publication, however this doesn’t always mean that the patches are available for devices from all vendors.”

Users can find their respective Android phone’s version number, security update level, and Google Play level in the Settings app. Generally, Android pushes automatic updates to users, but one can check for updates.

Quote:Journalist Matt Taibbi, who broke the sensational “Twitter Files” story that exposed the inner workings of the social media giant’s censorship machine, has announced he’s leaving Twitter in protest of apparent changes that have made the platform unusable for him.

Taibbi, who posts his articles on Substack and is one of the most popular contributors on the platform, made the announcement in a post titled “The Craziest Friday Ever” and a series of tweets, in which he said that he had just learned that Substack links were being blocked on Twitter.

“When I asked why, I was told it’s a dispute over the new Substack Notes platform,” Taibbi wrote in a tweet.

“It turns out Twitter is upset about the new Substack Notes feature, which they see as a hostile rival,” Taibbi wrote on the Substack platform, adding that when he asked how he was supposed to market his work, he was given the option of posting his articles on Twitter rather than on Substack.

“Not much suspense there; I’m staying at Substack,” Taibbi wrote. “Beginning early next week I’ll be using the new Substack Notes feature (to which you’ll all have access) instead of Twitter, a decision that apparently will come with a price as far as any future Twitter Files reports are concerned.

Are you sure, Mr. Taibbi?
You better read this article published on March 28th right away!

Quote:Substack is desperate, huh? That’s what I understand from their fundraising email, anyway. They’re now hitting up retail investors for millions of dollars after they failed to raise last year.

Substack Wrote:When we raised our last round of funding, in March 2021, we explored how we might make it possible for a large group of writers to invest alongside the traditional investors, but it ultimately proved too complex. Most importantly, it was difficult to include people who were not already accredited investors—a qualification determined largely by wealth. But the idea never left our minds.

Quote: Andreessen Horowitz led that round, which gave Substack a valuation of $650 million, and a16z has been merrily dumping on retail through their crypto investments for some time. It does not surprise me that someone might have thought Substack could expand the strategy!
...
You see, the last time Substack raised, the Fed hadn’t started its rate hikes yet. Startups — like Substack — are particularly vulnerable to being squeezed when the interest rates go up. It gets harder to raise money because conservative investors can simply invest in safer assets.

And during that 10-year period I cited with those outsize returns, interest rates were low and valuations of private companies ballooned. Now, with interest rates coming back up, those balloons are popping. Some VCs are slicing valuations by as much as 95 percent. There may be even more write-downs coming. And following the collapse of Silicon Valley Bank, there’s a considerable amount of uncertainty in the VC world.

Substack certainly knows this. It tried to raise last year, seeking $75 million to $100 million from investors. But it had revenue of only $9 million in 2021, and a sky-high valuation on relatively little revenue was not the vibe in 2022. The company gave up. On its Wefunder page, the company says that the pre-money valuation on Substack is now $585 million, a 10 percent decrease from 2021.

And now Substack has turned to Wefunder and retail investors. Friends, I do not like it, not least because the VCs last year got a pitch with Substack’s annual revenue, and I do not see that s*** line-itemed anywhere on the Wefunder page.

Substack makes its money by taking a 10 percent cut of the subscription fees its newsletter writers charge. (Its payment processor takes another 4 percent, according to Wefunder.) The company says it paid out more than $300 million to writers, cumulatively.

It seems that some terribly dark days loom in Substack's horizon, Mr. Taibbi.