Closing Bell: 10/15/26

Meta Platforms

Across the shows: 25 speakers in other episodes, 26 bullish and 6 bearish calls, past 30 days

• Mark Mahaney · Evercore ISINeutral

  Street expectations for Meta's next-year expenses and capex are more likely to fall than rise.

  Mahaney expects Meta to give next year's expense and capital expenditure guidance the following week. He thinks street numbers for these are more likely to come down than go up, while he expects Google and Amazon numbers to rise near term. This is part of the broader point that hyperscaler free cash flow visibility is the key overhang, since spending is currently high and the companies are raising a lot of capital.

  “When they give their guidance next week in terms of next year's expense and CapEx, capital expenditures, I think street numbers are more likely to come down than to go up.”at 50:41

• Mark Mahaney · Evercore ISIBullish

  Meta's Muse consumer AI app has taken the market; no freemium rival can challenge it.

  Mahaney says Muse changed the game and that the market had become fixated on enterprise AI while forgetting consumer AI. He quotes Zuckerberg that people need AI that works, not necessarily the most advanced models, and says Muse delivers that for the average person. He says the product is great and that Meta has effectively stolen the market, with no one coming in with a freemium product to compete. He says the stock has performed well and that Meta has already delivered its product-cycle catalyst, which is why it is no longer his pick.

  “Muse kinda changed the game... I think they just stole the market. It's all theirs. Nobody's coming in with a freemium product to take on Muse.”at 51:45

Anthropic

Across the shows: 39 speakers in other episodes, 20 bullish and 39 bearish calls, past 30 days

• Jeff Richards · Notable CapitalBullish

  Anthropic is a great long-term bet; public fears from leaked numbers are overblown and will ease once it publishes data.

  Richards argues the market is reacting to rumored or leaked Anthropic numbers rather than published data. He cites reported growth of about 10x year over year and net revenue retention reported above 900%, which he says would be among the highest metrics in IPO history. He expects that once Anthropic publishes numbers and analyst coverage arrives, investor fears will ease. He describes it as a long-term bet and says Notable is a long-term investor in the company, which Scott Wapner notes is an early investor.

  “I think as they come out and are publishing their numbers and you have analyst coverage and whatnot, this is such a great long term bet.”at 11:31

Waymo

Across the shows: 2 speakers in other episodes, 1 bullish and 1 bearish calls, past 30 days

• Jeff Richards · Notable CapitalBullish

  Waymo is one of the best examples of what AI will do for people; the benefit is becoming visible to consumers.

  Richards says he sees Waymo cars all over San Francisco and calls them incredible. He argues that AI driving with no human intervention could address the roughly 40,000 annual US road deaths, making Waymo a clear example of AI's end benefit. He notes the service is still limited to a few cities, but expects more consumers to see the benefit as it spreads, which he says could change the narrative around AI and data centers.

  “And as more and more people around the country, to me, Waymo is one of the best examples of what is AI going to do for us.”at 16:10

Amazon.com

Across the shows: 10 speakers in other episodes, 5 bullish and 3 bearish calls, past 30 days

• Mark Mahaney · Evercore ISIBullish

  Amazon is nearing a free cash flow inflection with accelerating AWS growth and high margins; a buy.

  Mahaney argues Amazon is approaching a free cash flow inflection, which he expects to show in the next quarter or two. He expects AWS revenue growth to accelerate sharply with high margins, driven by AI demand and falling costs. He says investors can buy Amazon at about 22 times earnings with less risk, and that the rest of the business is doing well. He contrasts this with hyperscalers that are still spending heavily, raising capital and not yet generating visible free cash flow, which is the overhang on the stock.

  “So my topic is Amazon, and I think you got a free cash flow inflection point coming up. They're gonna give you pretty sharply, I think, accelerating revenue growth at AWS with high margins.”at 49:58

Alphabet

Across the shows: 15 speakers in other episodes, 13 bullish and 5 bearish calls, past 30 days

• Mark Mahaney · Evercore ISIBullish

  Alphabet needs a Muse-rival launch as a catalyst; its near-term estimates should rise.

  Mahaney expects Alphabet's street numbers to go up in the near term, similar to Amazon, and sees a product-cycle catalyst for Google. He hopes Google will launch a competitor to Meta's Muse, because that launch would be the catalyst Alphabet needs given Meta has already had its catalysts. His view depends on Google delivering that launch, so the case is conditional.

  “Near term, I think it's the opposite with probably with Google, and I think very likely with Amazon that numbers go up... Google, also like. I think we have a catalyst there. I hope we better have a catalyst there, and that better be their launch of their Muse competitor.”at 50:41

AI: overall outlook

• Jeff Richards · Notable Capital

  AI is only in the first inning; its long-term trajectory could mirror the AWS-era cloud build-out.

  Richards compares AI to Amazon's launch of AWS around 2005-2006. In 2010 UBS forecast a cloud addressable market of about $5 billion a year when Amazon was doing about $500 million. Today the hyperscalers combined generate about $350 billion a year. He argues AI is at a similar early stage, so the long-term trajectory matters more than near-term volatility, even with inflation, a war in Iran and rising rates in the background.

  “So I think with AI, you're looking at something similar where we're super early. We're in the first out of the first inning.”at 12:16

• Nikesh Arora · Palo Alto Networks

  The public has over-rotated to AI's negative effects; the narrative should shift to its large benefits.

  Arora says AI's benefits are huge if harnessed and deployed well, and that technology has never before over-rotated this hard onto negative effects. He cites drug discovery and the democratisation of intelligence, saying people now carry a super IQ in their pocket. He ties public fear to headlines about hacked bottles and agents attacking Hugging Face, and says the industry must convince people that AI will create jobs, lift people out of poverty and improve society, or it will be seen as something to avoid.

  “that we have over rotated towards the negative effects of it. It's never happened in technology before. We've never over rotated so hard onto the negative effects.”at 37:21

• Mark Mahaney · Evercore ISI

  AI demand is early; enterprise and consumer adoption have barely started, so much demand is still to come.

  Mahaney says the market is waking up to the fact that AI demand has a long way to go. He argues enterprise adoption is still early and consumer adoption even more so, so the company has not yet reached the first inning. He ties this to demand that is still to come, which is the basis for his view that AI-linked stocks have room to run.

  “I think the market is waking up to a couple of things. One is that AI demand, we're still got so far to go. We're still so early on in terms of enterprise and especially consumer adoption. We're not even... We haven't even started the first inning.”at 49:19

Enterprise AI adoption & AI software

• Jeff Richards · Notable Capital

  AI adoption is still early: most small businesses use it only primitively, and demand is growing fast.

  Richards says more than 70% of small businesses now use AI, but in a primitive way, such as searches or payroll optimization. He expects usage to ramp up and notes that small business owners are also consumers. He adds that infrastructure-layer companies such as Databricks, Fireworks, Baseten, Vercel, Modal and Fal are being relied on and bought by large enterprises. He says these companies are growing 60% to 100% a year, and once they publish metrics the demand side will become clearer.

  “Small businesses today, more than 70% are using AI. Now they're using it in a very primitive way, probably for searches or maybe helping them optimize payroll or whatnot.”at 17:11

• Daron Acemoglu · Massachusetts Institute of Technology

  Enterprise AI diffusion will be slower than proponents predict, because models lack judgment for many jobs.

  Acemoglu argues that slow AI diffusion means slow job and productivity effects. He says models lack judgment and are optimized for narrow metrics, so when they escape sandboxes they do unforeseeable things. That makes them unsuitable for air traffic control, CEO work or even customer service right now. He says the industry has not invested enough in applications. He hedges that he would not rule out productivity gains from AI over the next ten years, based on what has happened so far.

  “I don't think the spread of AI is going to be as fast as some of its biggest proponents are predicting.”at 22:29

• Nikesh Arora · Palo Alto Networks

  Enterprises will adopt AI only as fast as they trust it, so risk must be managed to speed deployment.

  Arora says enterprises will not deploy AI if they believe there is an element of risk, so AI will be embraced only at the pace enterprises feel it can be trusted. He concedes that frontier models are very powerful and could help bad actors find vulnerabilities. He argues the response is to understand how such attacks happen and prevent them, not to accept defeat. Trust is what allows enterprises to adopt the technology.

  “AI will be embraced at the pace enterprises feel comfortable that it can be trusted.”at 29:53

AI regulation & safety

• Nikesh Arora · Palo Alto Networks

  Frontier labs should build hooks so third-party security firms can observe AI agents, not just fear-monger.

  Arora says AI labs should stop fear-mongering and expose the tools and hooks in their products that let third-party security companies observe what the models do. He notes technology has historically had a second set of eyes from the security industry. Observability builds trust with enterprises so AI can spread more robustly and securely. He also argues that AI decisions need explainability, since a company cannot tell a mortgage applicant that AI did it. He wants to see why AI does something while it is happening, so systems can intercept it, through an AI or human kill switch.

  “And I think part of what needs to happen is that the AI labs need to get off this fear mongering business.”at 33:10

• Nikesh Arora · Palo Alto Networks

  Existing consumer and enterprise regulations are enough; no new federal AI rules are needed.

  Asked whether federal guardrails beyond current rules are needed, Arora says there is enough regulation already in place to protect consumers and enterprises. He argues that existing compliance and explainability requirements, such as those for loan decisions, already force companies to understand their systems. This is his view that the gap is in implementation and observability rather than new law.

  “I think there is enough regulations that exist that are there to protect consumers. There is enough regulations that are there to protect enterprises.”at 35:38

• Nikesh Arora · Palo Alto Networks

  Enough regulation already exists; the industry should self-organize on safe deployment, observability and guardrails.

  Arora argues there are already enough regulations protecting consumers and enterprises. He says what is needed is for the whole industry to rally after the recent White House initiative on diffusing the technology safely and securely, rather than treating it as a research project, because AI is becoming mainstream and will be used by almost every company and consumer. He says agents that move money need observability, built-in security and guardrails that third parties can check to confirm models behave as advertised.

  “There is enough regulation out there. I think what we need is... We need the entire industry to rally together in terms of trying to understand how to diffuse this technology in a safe and secure manner, as opposed to treat it like a research project”at 35:46

IPOs, M&A & deal activity

• Jeff Richards · Notable Capital

  Failed AI-era IPOs show healthy market discernment, not weakness in the AI IPO pipeline.

  Richards argues that the failed or withdrawn IPOs of Xelevation, Firmus and Aura have no real correlation with the AI IPO market for strong companies. He says the market pushing back is a healthy sign of discernment. Firmus, a $50 million revenue company seeking a $30 billion valuation with a manager who had served jail time for insider trading, was not a good bellwether. He contrasts this with the 2021 SPAC era, which he sees as a lack of discernment. He says SpaceX's strong IPO, the biggest of the year, is a better barometer of demand for high-growth names, which bodes well for Anthropic and OpenAI.

  “I think it's a sign of a healthy market when the market's discerning.”at 13:52

AI & jobs

• Daron Acemoglu · Massachusetts Institute of Technology

  AI is built as automation that will take many jobs; without pro-worker design, displacement could damage society.

  Acemoglu argues that the industry is developing AI as an automation technology under the banner of AGI, designed to replace humans, so it will eventually take many jobs. He says jobs created by AI are either transitory, such as building data centers or training models, or incidental, and do not offset the displacement. He says the leading labs are not pursuing pro-worker AI, which would make workers more productive. He warns that if 30% to 40% of graduates cannot find work, democracy and social peace could suffer. He argues the job losses leaders project are much larger than those from China trade and robots. His predictions would be more positive if the agenda changed.

  “And the way we are developing AI right now is as an automation technology, which means that it will ultimately be capable enough to take a lot of jobs.”at 21:48

Frontier AI labs & model race

• Daron Acemoglu · Massachusetts Institute of Technology

  Consumer chatbot subscriptions cannot cover AI labs' revenue projections; they need business automation revenue.

  Acemoglu looks at the revenue projections of the AI industry and argues they cannot be reached from consumer chatbots alone. He says the labs need to make a lot of money from businesses, and that will come from automation. This makes the business case for automating work central to the industry's economics, which is why he thinks the labs will push for automation even as adoption is slow.

  “But if you look at the revenue projections of the industry, there's no way that you're gonna be able to generate the kind of revenues that they need just from chatbots sold to consumers.”at 24:12

Cybersecurity

• Nikesh Arora · Palo Alto Networks

  AI-driven attacks outpace current response times; defenders must move to real-time response to close the gap.

  Arora argues cyber attacks have run at a steady pace for years, and AI is mainly making vulnerability discovery and movement faster. Defenders must increase the speed of defense to match offense, patch vulnerabilities and find them before bad actors do. He says the biggest change needed is a real-time response. AI could get in and out of infrastructure in under twelve minutes, while the industry's response takes about four days, which he calls asymmetric. He agrees with Jensen Huang that the industry should not accept that incidents will happen.

  “So you have to increase the speed of defense just the way the speed of offense is increasing... the biggest thing that needs to change is we have to make sure that our response time gets faster.”at 31:21