Farzad: "Even the Data Center Issue Is Short Lived" + "Tesla and SpaceX FAR Too Valuable to America"

The day after Tuesday's elections, the data center buildout was the big topic in Farzad's feed. Wednesday at 11:29am CT he wrote that "even the data center issue is short lived." Jay Yoon expects Texas delays, not a shutdown, and Daniel Newman writes that "location is now fungible." Farzad also called a predicted Tesla and SpaceX market reckoning "a bunch of bullshit."

"Even the data center issue is short lived"

Farzad, Wednesday at 11:29am CT, in a reply: Even the data center issue is short lived. AI is too additive to the economy.

He goes on: "Slowing down AI means slowing down the economy." He expects "a couple projects might be screwed as sacrificial lambs," but says "the reality is that the growth will continue to accelerate because the demand is FAR too big, and the technology is FAR TOO IMPORTANT for national security and defense."

That's his projection. He doesn't name the projects or give a timeline.

amit (amitisinvesting) replied at 11:31am CT that it "might be a perception thing," and that "too much of it is most of our capex." His read: "if anything they are going to work very closely with the tech companies to properly get these datacenters up and running." He doesn't say who "they" are. Farzad liked the reply.

Earlier, at 10:30am CT, Farzad quoted a 9:07am CT post from Andrew Curran about Tuesday's local votes on datacenter moratoriums and restrictions. Curran wrote that public sentiment on AI in America "has been going steadily downhill for three years" and "is once again at a new low."

Farzad's take: "Companies will find states/municipalities that are willing to work with them." He says they'll "need to give more carrots," like funding schools and projects and paying for people's electricity and water, "while ensuring water/noise/pollution not an issue." And: "AI satellites gonna very obviously boom too."

Check Farzad's reply here. amit here. Andrew Curran here. Farzad's quote post here.

Texas: delays, not a torpedo

Jay Yoon, Wednesday at 11:29am CT: Texas won't torpedo its datacenter industry, but it will introduce some delays while it works through finalizing legislation to make datacenters more palatable for its residents.

He says this "would mostly affect capacity scheduled to come online in 2028 versus 2027." Because "datacenter location is fungible to a degree," he writes, "CSPs and neoclouds can pivot to other regions with less opposition / permitting delays." His estimate: it "could push out maybe 5-10% of total 2028 datacenter capacity scheduled to come online in 2028 by 2-3 quarters, which is hardly catastrophic."

His warning is aimed at AI labs. If Anthropic and OpenAI "keep up with their braindead messaging of AI will take all your jobs, hack all of your bank accounts, and eventually send a T-1000 to your house to kill your entire family," he writes, "then yeah we might see more widespread pushback and a catastrophic slowdown." He closes: "Hopefully they aren't that stupid."

Those numbers are Jay's estimate. He doesn't name the legislation or give a source for the 5-10%. Farzad liked the post and quoted it at 2:55pm CT: Yes.

A minute earlier he'd quoted 0xDevShah with the same word. 0xDevShah, 11:10am CT: Opposing data centers is not going to make the demand for compute go away.

The post adds that data centers "will bring a plethora of local blue-collar jobs in construction, energy, infra, and the industries that grow around them." It says big tech "needs to be far more conscientious of the towns and states making this possible," and closes: "This should be a positive-sum game for everyone."

Check Jay Yoon here. Farzad's quote post here. 0xDevShah here. Farzad's quote post here.

"Location is now fungible"

Daniel Newman, Wednesday at 1:48pm CT: Spoke to a few large semiconductor and infrastructure CEOs and consensus is if the build slows in the U.S. the demand outside the U.S. is more than sufficient.

He adds that forecasts remain intact, and then: "Location is now fungible." He doesn't name the CEOs.

Farzad quoted it at 2:52pm CT: Hey @grok which states in the US are friendliest towards Data Center buildouts?

Grok's answer isn't in this pull.

Jeff Lutz, Thursday at 6:45am CT: Reminder that federal land can and is being used for AI datacenter deployments. Super intelligence is a national security issue

It includes an image this pull doesn't describe, and the post doesn't name any sites. Farzad liked it and RTd it at 8:37am CT.

Farzad RTd and liked two more Wednesday posts. Kevin O'Leary, 2:24pm CT: The politics around AI data centers are about to change fast.

He writes that states "may have pushed back on these projects before," but once they see "the thousands of high paying jobs, new power infrastructure, and capital investment attached to them, that resistance is going to disappear." Farzad RTd it at 4:08pm CT.

Nathaniel Whittemore, 2:45pm CT: I don't think data centers stop getting built.

He expects leaders to demand "No NDAs/Full process transparency," builders "Paying full cost of infra/electric (and subsidizing everyone else's)," "Actual consideration for quality of life concerns" and "Big $$ involvement in local communities." His call: "the builders will positively JUMP to agree to it all," and the system keeps turning "in a way that's frankly better and more sustainable." Farzad RTd it at 3:08pm CT.

Check Daniel Newman here. Farzad's question here. Jeff Lutz here. Farzad's RT here. Kevin O'Leary here. Farzad's RT here. Nathaniel Whittemore here. Farzad's RT here.

A bull case and a "fast AI takeoff"

Ejaaz (cryptopunk7213) started Wednesday cautious. At 9:53am CT he wrote: "man people really don't like AI." He expects this to "slow down a lot of momentum in the near-term (6-9 months)," and says "around $400-750 billion worth of data center projects are now at risk of being shut down or paused next year." The post says 25% of that is in Texas and that "76% of people don't want them built in their community." It ends: "the doomerism has won."

At 5:09pm CT he quoted himself with what he calls the AI bull case, which he says he believes "will play out." His points:

Data centers "will not get cancelled, they'll be taxed more."

"no ai bubble burst because data center buildout slows. can't overbuild if you can't… build."

Every AI lab, hyperscaler and neocloud agrees to state leaders' requests "to remove subsidies, increase taxes blah blah (they make so much money they don't care)."

The stock market "recovers dramatically" because "the long term outlook of AI's impact on the economy is still massively undersized."

That's his opinion. Farzad liked the 5:09pm post.

gfodor, 11:17am CT, says the commentary he's seeing leaves out "that we are in a hard, pretty fast AI takeoff." He writes: "Absent a civil or global war, everything Kurzweil predicted is going to happen." He says "we have automated frontier mathematics, have solved vision, creativity, and seemingly continual learning," and calls the implications "purely inductive: massive deflation, mass production of humanoid robots, and widespread human augmentation." He adds: "The average IQ of humans is going to rise over time going forward." Those are his claims, and he cites no sources for them. Farzad liked it.

Check Ejaaz's first post here. His bull case here. gfodor here.

"FAR too valuable to America"

Jamie Metzl, Wednesday at 7:04am CT: Tesla and SpaceX will face a severe market reckoning today. Investors should expect a rapid drop in the share prices of both companies.

The rest of his post criticizes Elon's role in politics and predicts congressional hearings with subpoena power. That's Metzl's opinion.

Farzad quoted it at 9:46am CT: I think this is a bunch of bullshit.

And: Tesla and SpaceX FAR too valuable to America.

That's Farzad's opinion against Metzl's prediction. This pull doesn't show how either stock traded. At pull time Farzad's post had 621 likes, and Metzl's had 1,030 replies.

Check Jamie Metzl here. Farzad's quote post here.

"Only victory for steering wheel and pedals"

Kris Patel, Wednesday at 6:37am CT: Now might be a good time to go long $UBER

He argues that "competitors of Musk companies" could "get rerated higher" over the next few years and says Uber "is one of the largest employers in the US." The rest argues autonomous taxi companies could face a more hostile political environment, and says the same goes for SpaceX.

Farzad replied at 11:48am CT: If that's the case Tesla will simply put steering wheel/pedals in Cybercab and sell millions more cars per year.

Two minutes later, at 11:50am CT, he posted:

Dear Tesla community,

Do you hear that in the distance?

Do you feel the energy in the air?

And you really thought you got rid of me.

Yet everything is going LITERALLY how I described it.

Too late for apologies.

Only victory for steering wheel and pedals.

tesla_lion asked at 11:57am CT whether this was "farzad's grokbot or farzad speaking." Farzad, 12:13pm CT: No.

CernBasher, 12:48pm CT: "At this point I don't think anyone really knows what you think you're right about?! Consider clearly restating it." Farzad, a minute later: You know exactly what I'm talking about.

Kris pushed back at 1:30pm CT: "The point of autonomous was the massive margin increase from not having a human at the wheel? What's the value proposition now if a human is required for every ride?"

Farzad, 1:33pm CT: Because Unsupervised FSD is still something Tesla can sell on a recurring revenue basis where people don't have to pay attention - they just need to be in the driver's seat in case something goes wrong.

And: That's a huge profitability unlock.

All of this is Farzad's projection. As built, the Cybercab has no steering wheel or pedals, per Reuters. None of these posts cites a Tesla announcement, and he doesn't restate in this thread what he says he predicted.

Check Kris Patel here. Farzad's reply here. "Dear Tesla community" here. tesla_lion here. Farzad's reply here. CernBasher here. Farzad's reply here. Kris's pushback here. Farzad on Unsupervised FSD here.

"MASSIVE blue collar hiring"

Farzad, Wednesday at 1:48pm CT: My prediction for 2027.

His five lines: "MASSIVE white collar layoffs. MASSIVE blue collar hiring. MASSIVE corporate earnings. MASSIVE AI profitability. MASSIVE growth in new companies."

That's his projection. carmineglitch replied that Southern California "still has massive demand for Electricians" and that the work opening up "is 3-5 years of constant work for alot of new hires." Farzad liked the reply.

bgurley, 1:56pm CT, wrote that "AI will take 50% of jobs" and "AI will cure cancer" are both overstatements. Farzad replied at 2:04pm CT: "I think AI causing a lot of white collar layoffs is inevitable." His reason: corporations "are typically extremely bloated," and AI "becomes a super easy way for companies to massively reduce expenses."

Earlier, at 1:38pm CT, he wrote: "If AI build out slows down, token prices go UP because there will be FAR more demand vs supply." He says that "locks AI only for the people that can afford it." His fix: "The only way to reverse this is by building FASTER!" Matt Farley replied that he's "Bullish space launch providers too," because companies will likely "take to space if regulators shut down new terrestrial data centers." Farzad liked it.

At 11:27am CT, Cyber_Trailer asked whether closing the wealth gap meant confiscation. Farzad's answer: "You don't have to confiscate. Divert significant % of AI profits to fund basic income for citizens." And: "Basically the Alaska oil fund on steroids."

Check Farzad's prediction here. carmineglitch here. bgurley here. Farzad's reply here. Farzad on token prices here. Matt Farley here. Cyber_Trailer here. Farzad's reply here.

Lemonade Q3 2026

Shai Wininger, Thursday at 5:03am CT: 12 straight quarters of accelerating growth. @Lemonade_Inc's Q3 2026 results are in!

His year-over-year highlights: IFP reached $1.55 billion, up 34%, the 12th straight quarter of accelerating IFP growth. Revenue grew 73% to $336M. Gross profit hit a record $118M, up 48%. Customers reached 3.52M, up 23%. Adjusted free cash flow was $27M, the 6th straight positive quarter. Cash and investments came to $1.2 billion.

On Car, he says Car IFP grew 63%. It's now in 14 states, "covering more than half the U.S. population," and "Autonomous Car for Tesla FSD now in 6 states." He doesn't name the six.

Under Lemonade OS he lists more than 2.5x the premium base "with fewer employees than four years ago," 25 new product-geography combinations this year, Renters now nationwide, "Native support for personal AI agents coming soon," and live demos at Investor Day on Nov 17.

On guidance, he says Lemonade expects its "first-ever positive adj. EBITDA quarter in Q4," reiterates its expectation of "a full year of positive adj. EBITDA in 2027," and is raising full-year revenue and gross earned premium guidance.

Lemonade's Q3 2026 shareholder letter matches the headline numbers: IFP of $1.55 billion, up 34%, revenue of $336 million, up 73%, and record gross profit of $118 million. It also reports a net loss of $44 million. Farzad liked the post. Disclosure: Farzad owns Lemonade stock.

Check Shai Wininger here.

Farzad's videos

Farzad posted four videos between Wednesday at 12:15pm CT and 9:06pm CT. This pull has only their titles. Under three of them he replied "Build your own short form content with my Grok Bot Shorty." Under the SpaceX video he replied "Made with my Grok Bot Solo." Farzad also liked several viewer replies praising the SpaceX video.

Check Why Tesla's Giga Press Needs 6,000 Tonnes of Force, Wednesday at 12:15pm CT, here.

Check SpaceX Just Made A Shocking Announcement, Wednesday at 12:20pm CT, here.

Check Why the James Webb Telescope Hides Behind a Tennis-Court Sunshield, Wednesday at 5:58pm CT, here.

Check Why Weather and TV Satellites Sit Exactly 36,000 km Up, Wednesday at 9:06pm CT, here.

Check the Solo reply under the SpaceX video here.