The Great Slopification, Powered by OpenAI

OpenAI just unveiled Sora 2: its latest audio and video generator. The tool, which is invite-only for now, lets users create digital cameos of themselves and others in videos and even generates speech. 

OpenAI also launched a new social platform called Sora for sharing and discovering AI-made clips, positioning itself as a competitor to TikTok. Meanwhile, Meta is rolling out its own version, called Vibes, also focused on short-form, AI-generated clips. 

Some people are very excited about this, while others are calling these feeds the next iteration of “AI slop.”

The bigger issue isn’t whether the AI-generated art is “good” or “bad.” It’s that most consumers don’t actually want to create it in the first place. Media consumption has long followed the “1% rule”: Only a small fraction of people create content, while the vast majority consume it.

History suggests that consumers don’t want to produce their own entertainment. They want to be entertained by a small group of creators who are especially good at telling stories.

Ever since ChatGPT image generation was released in 2022, Google search interest for AI image generation has fallen nearly 80%. Another study showed AI art in Reddit communities has plummeted.

OpenAI’s bigger opportunities likely lie in high-margin businesses like B2B SaaS, digital advertising, and payment facilitation (pay with ChatGPT). 

A key question is whether OpenAI’s massive funding base reduces discipline. With seemingly unlimited capital, the pressure to prioritize is lower. The risk is that the more time OpenAI spends on moonshots, the more room competitors have to secure customers in higher-margin sectors — such as Anthropic in B2B. 

When the first iteration of Sora launched last year, OpenAI presented it as a tool for the creative industry. They solicited feedback from artists and filmmakers “on how to advance the model to be most helpful for creative professionals.”

In Sora 2’s release, the OpenAI team framed Sora as a tool for connecting with friends and family. From the release video: “Social media has moved away from the idea of friends and family connections. … Sora can lean into this because it’s just so easy to create.” 

What this tool is really about isn’t connection – it's about letting studios and advertising agencies dramatically reduce the cost of producing compelling images and videos, i.e., fire people. 

OpenAI knows Sora’s potential in Hollywood. Sam Altman has already been in LA pitching Sora to entertainment execs. 

The reason they’re burying the lede is because being up front about the impacts of their revolutionary technology earned them bad press last time. 

The bottom line is that this is clearly a B2B tool being masqueraded  as a B2C “connection” mechanism.

You would have thought a year ago when these apps came out that the first to go would be poor design firms and poor designers. But what you’ve seen happen at tech companies is the ratio of designers to programmers has actually gone up because as everyone starts to use AI to make their websites and brand assets, one to stand out is to hire humans to make great UX and differentiated design.

OpenAI's hiring designers. OpenAI even paid nearly $7 billion for Jony Ive to cosplay his younger self. God, that was dumb. The video of him and Sam Altman was like when Harry met Sally — “I'm just a billionaire looking at another billionaire asking him to make me a trillionaire.” 

Global M&A activity is rebounding … sort of. Global deal values hit $1 trillion in a third quarter for only the second time ever. But the headline numbers obscure a concerning trend. While deal value increased 40%, deal volume declined 16%. Translation: Big deals are back, but small ones are vanishing. In the first half of the year:

The reason: Only the largest firms are still playing offense. Small and midsize companies have a hard time during macro volatility and political instability, while large corporations are better equipped to handle shifting supply chains or absorb temporary shocks.

Who is a big winner in all of this? Investment banks. Investment banks have generated $95 billion in fees this year, the second-highest year to date (YTD) ever. Bank stocks are ripping:

I just find it remarkable how there’s one consistent trend running through every story we talk about, and that’s inequality.

We see it in the consumer economy. We see it in the stock market. I mean, the fact that the Magnificent Seven now makes up 35% of the entire market cap of the S&P right now. In 2019, it was 19%. So you’ve gone from a fifth to a third in just about five years. And the downstream effects of that inequality are spreading out everywhere, including M&A.

When you dig into the numbers, there’s been a ton of deal-making among the large caps, and then the opposite is happening among small-cap companies. Deals worth less than half a billion dollars are down 18% this year, but deals worth more than $10 billion are up 26%. The small companies are not really being touched when it comes to these big acquisitions. 

And what does that mean? That means fewer startup exits, fewer incentives for entrepreneurs, less innovation, and fewer public companies overall.

Last week, Charlie Javice was sentenced to seven years in prison for defrauding JPMorgan. Back in 2021, she convinced the bank to pay $175 million for her startup, Frank, by claiming it had 4 million users. In reality, the platform never had more than 300,000.

This case raises a bigger question: Where does ambitious, visionary storytelling become outright fraud? Elizabeth Holmes, Sam Bankman-Fried, and Trevor Milton all crossed that line. Now, there are signs that AI founders are making the same mistake.

According to a report published by Fortune, AI investors are starting to worry about “creative accounting” issues, where companies are signing one-off contracts and then annualizing those numbers as “ARR” so that it seems like they’re making more money than they really are. As one investor told Fortune: “The problem is that so much of this is essentially vibe revenue.”

In some cases, the accounting is downright fraudulent. 11x, a startup valued at $700 million, reported $14 million in ARR. Turns out that wasn’t true. They were counting canceled contracts in their ARR number — in reality, revenue was closer to $3 million.

Rezolve AI went public via SPAC merger in late 2024 claiming to be an AI company. However, according to former employees, it was all marketing hype: The company had 0 revenue from AI in 2024. 

AI hype has created a feedback loop where startups stretch ARR, boosting valuations and raising new investor interest and FOMO-based funding. Median late-stage AI startup valuations are already 3x higher than valuations of non-AI companies. No doubt there’s real innovation here, but what happens if the foundation of this new tech cycle rests on circular revenues and “vibe-based” ARR?

If you’re the CEO of a startup, you have to be able to spin a narrative that gets people’s greed glands going. And sometimes there’s a thin line between exaggeration and downright lying. But what you should never, ever lie about is the numbers. The number of customer accounts, the revenues. 

I also think that at the end of the day, with your board, you wanna underpromise and overdeliver. You especially want to overcommunicate when things are bad. 

If you lose your biggest client at 4 p.m., you want to put an email out to your board at 4:01 saying, we lost our biggest client. When they don’t hear from you, that should be because everything’s going great. Boards don’t mind bad news. What they mind is surprises. “Wait, this happened two months ago and we’re just finding out about it now?” That’s when you get fired and they get upset.

Podcasts have become a primary way we learn about AI and tech. Increasingly, the people behind the mic aren’t journalists or academics but venture capitalists. One in five of the top 50 tech podcasts are VC-hosted: All-In (Jason Calacanis, David Sacks, and others) tops the charts, while shows from Harry Stebbings (20VC), Jack Altman (Uncapped), and Andreessen Horowitz dominate the space.

This matters because VCs aren’t neutral interviewers. They often invite founders they’ve backed — or hope to back — using the podcast as both PR for portfolio companies and a funnel for new deals. Calacanis even told investors his show was a key driver of startup sourcing for the fund: “The podcasts are the top response we receive when asking founders how they found out about our firm,” his fundraising materials read.

The impact is that these podcasts rarely challenge founders. Instead, they amplify hype, reward magical thinking, and set the narrative in ways that serve investors’ interests — not listeners seeking clear-eyed analysis of AI’s risks and trade-offs.

(Note: While he isn’t a VC, Scott is a private investor. For transparency, Scott discloses his interest in companies that are discussed on his podcasts and newsletters.)

The most boring member of the Magnificent Seven in 2025 has been Amazon. It’s also been the worst performer. It hasn’t captured the spotlight with AI ads like Meta, translation breakthroughs like Apple, or autonomous driving promises like Tesla.

But in business, boring is sexy. 

Everyone points to AWS as Amazon’s growth engine, but its overlooked, and boring, long-term advantage lies in industrial robotics.

Amazon Robotics is the largest warehouse robotics company in the world — and, outside of Tesla, it’s the only U.S. firm meaningfully competing with China’s dominance in the sector. 

Amazon already operates over 1 million robots across its warehouses, nearly matching its human workforce. Since 2022, it has rolled out six new robot models and added a new AI-based robot traffic management system, DeepFleet. These technologies are expected to boost productivity by roughly 25% at Amazon’s fulfillment centers.

Policy is also a tail wind. The 2025 tax reform bill made 100% bonus depreciation permanent for machinery, robotics, and automation equipment. Instead of its being phased out under the old Tax Cuts and Jobs Act, manufacturers can now expense capital investments immediately.

Historically, Amazon’s margin expansion has been powered by ads and AWS. Now retail is contributing. For over a decade, fulfillment and shipping costs ballooned more quickly than sales. That reversed two years ago; retail sales are now growing faster than shipping costs, suggesting robotics is finally delivering operational leverage. 

$10 billion in cost savings, based on last year’s financials, translates to an additional $170 billion in enterprise value.

Amazon’s robotics won’t just yield cost savings; they’ll also strengthen Amazon’s most durable moat — fulfillment infrastructure. ChatGPT may change how people search and shop, but it can’t replicate the logistics network that gets cough syrup from a warehouse to your door in six hours.

Despite all of this, Amazon’s valuation looks low relative to history. Shares trade at 34x earnings, well below the five-year average of 60x. The market is pricing in AWS’ dominance but misses the retail margin story. That makes Amazon one of the most underappreciated members of the Magnificent Seven heading into 2026.

Depreciation is an accounting practice where the purchase price of a major business asset (such as equipment, structures, or vehicles) is gradually written off over the period the asset is expected to be useful. This process serves a few key financial purposes:

Bonus depreciation is an accelerated depreciation method that allows a business to deduct a larger-than-normal percentage of the asset's cost in the first year. The goal is to give businesses immediate tax savings, motivating them to make new capital investments immediately.

Netflix is going to make a tectonic acquisition. They’re facing an existential threat: We’re raising a generation of consumers who can’t sit still for 90 minutes to watch a movie. People open TikTok or YouTube for short entertainment hits. And when they do turn on the TV, Netflix is just another tile. They now need more than just content to sustain their scale.

On Raging Moderates, Scott and Jessica Tarlov break down who really wins from a government shutdown, Trump’s surprising new peace plan in Gaza, and why MAGA is melting down over the NFL tapping Bad Bunny for the Super Bowl halftime show. Check it out here.