• Consider a diversified basket of power-generation and energy-infrastructure companies to capture rising data-center electricity demand, rather than betting on one technology; no allocation or price targets were provided.
• Vertiv (VRT) showed strong momentum, with revenue up 61% and orders more than doubling, but assess valuation before buying because no price target was discussed.
• GE Vernova (GEV) merits consideration for its $197 billion backlog and raised guidance, while its share-price decline underscores that strong results do not guarantee gains.
• Treat Bloom Energy (BE) as a higher-risk opportunity: wait for evidence of reliable deployments at scale given its high valuation and negative cash flow.
• For Amazon (AMZN), keep any long-term AI/cloud exposure sized prudently and avoid excessive leverage around earnings, when capital-spending concerns could drive volatility.

AI Data-Center Power and Energy Infrastructure

• The episode’s central investment theme is rising electricity demand from AI data centers. The hosts cited a Morgan Stanley estimate of a 38-gigawatt global AI power shortfall through 2028.
• They argued that major technology companies are signing long-term power-purchase agreements because they expect demand for electricity to persist. The discussion framed data-center power as a potential multi-year opportunity, while noting that the mix of supply could include natural gas, nuclear, and other sources.
• The hosts favored considering a basket of energy companies rather than betting on one technology or provider.
• Risks discussed include uncertainty about which energy technologies will win, the time required to bring new nuclear capacity online, and the possibility that AI-related power demand or investment expectations do not develop as anticipated.

Takeaways

• The discussion points to data-center electricity demand as a potential investment theme, but not a guarantee that every power supplier will benefit.
• A diversified approach across different power-generation models may reduce dependence on any one technology. The transcript did not provide a specific allocation or price target.

Vertiv (VRT)

• Vertiv reported 61% revenue growth and orders that more than doubled. Its stock rose 16% after the results.
• The company was presented as one of the businesses benefiting from the build-out of AI data centers and the infrastructure needed to support them.

Takeaways

• Vertiv’s reported growth and order momentum were positive signs in the AI infrastructure theme.
• The transcript did not discuss valuation, specific risks, or a price target, so the reported results alone do not establish whether the stock is attractively priced.

GE Vernova (GEV)

• GE Vernova reported $29 billion in orders and a $197 billion backlog, and raised its guidance. Despite those results, its stock fell.
• The hosts described GE Vernova as a supplier of turbines and noted that some of its business is booked well into the future, with sales extending to 2032.

Takeaways

• The backlog and raised guidance suggest substantial demand, but the stock’s decline illustrates that strong results do not guarantee a positive market reaction.
• The discussion presented GE Vernova as one possible component of a broader energy-infrastructure basket, without giving a price target or specific recommendation.

Bloom Energy (BE)

• Bloom Energy was described as a potential major beneficiary if AI data-center construction and electricity demand continue. The hosts said it has substantial demand and is beginning to scale after years of research and development.
• Its stock had fallen 20% in two days the previous week, then gained 18% over three days.
• The hosts highlighted potential advantages for data centers: Bloom systems can use a natural-gas pipeline, avoid drawing directly from the grid, and may be easier to deploy than some alternatives. They said the technology is cleaner than conventional natural-gas generation, but still relies on natural gas.
• They also noted that Bloom’s energy can cost more than alternatives, though they said costs have been declining.
• Risks discussed included Bloom’s high valuation, negative cash flow, and the need to prove that its systems can be deployed reliably at scale. The hosts said deployment problems at a major customer could prompt them to reassess the investment.
• A shortage of scandium was raised as a criticism in past short reports; the hosts disputed that concern, saying those reports did not account for improved efficiency and technology.

Takeaways

• The potential case for Bloom depends on continued data-center demand, lower costs, and successful large-scale deployments.
• The transcript’s main caution is execution: investors may want evidence that the technology works reliably at scale before assuming projected growth will materialize.
• One host said they owned Bloom; that personal position is not a recommendation.

Constellation Energy (CEG)

• Constellation was described as a power-generation company with a role in supplying electricity to large technology companies through long-term power-purchase agreements.
• The hosts said technology companies including Amazon, Meta, and Microsoft are signing long-term power agreements, and discussed nuclear power as part of the effort to meet demand.
• Constellation was included in the hosts’ proposed energy-stock basket.

Takeaways

• The investment case discussed is exposure to long-term electricity demand, including demand from data centers.
• The transcript did not provide Constellation-specific financial results, a price target, or a detailed risk assessment.

Vistra (VST)

• Vistra was described as combining traditional power generation, including natural gas, with efforts to expand nuclear generation.
• One host cited an expectation of roughly 20%–30% earnings growth for 2027 and said Vistra had recorded around 40% earnings growth that year. The transcript did not clarify the basis or period for those figures.
• Vistra was presented as a relatively straightforward way to gain exposure to power demand and was included in the hosts’ energy-stock basket.

Takeaways

• The discussion’s case for Vistra rests on electricity demand and a mix of existing generation and nuclear capacity.
• The growth figures cited are not guarantees; the transcript offered no price target or detailed company-specific risk analysis.

Amazon (AMZN)

• One host described having a long position in Amazon, which had grown in size and included some leverage and options. The host said they were cautious about adding substantial short-term leverage around earnings.
• The bullish view included Amazon Web Services’ role in AI infrastructure and the possibility that the company will benefit from ongoing AI investment.
• Risks discussed included investor concern about rising capital expenditures. Amazon’s cited full-year 2026 spending figure was $220 billion; the possibility of spending $300 billion–$350 billion the following year was presented as a scenario, not company guidance.
• Other concerns included higher diesel and energy costs affecting Amazon’s retail and logistics business, and higher memory costs affecting devices.
• The hosts also discussed the possibility that AI shopping agents could weaken Amazon’s advertising business. They said this could take years to unfold and noted that Amazon may develop its own agents or work with outside ones. They also cited concerns about fraud, returns, chargebacks, and customer control if third-party bots shop on the platform.

Takeaways

• The discussion presents Amazon as a long-term AI and cloud opportunity, but with potential near-term share-price volatility around capital-spending expectations.
• The hosts emphasized avoiding excessive leverage when market reactions are unpredictable. That caution is especially relevant around earnings and other events that could trigger sharp moves.
• The transcript did not give a price target.

Alphabet / Google (GOOGL, GOOG)

• The hosts argued that Google’s AI Mode could reduce the amount of search activity that is monetized as traditional search advertising. They said search historically accounted for about 85% of Google’s profits, while noting that AI-related activity may be difficult to monetize as effectively.
• They suggested that Google’s AI infrastructure, cloud business, and TPU chips could offset some pressure on search, but said they had not yet seen AI Mode monetized as effectively as traditional search.
• One host said they were not invested in Google and had previously held a short position, which they exited because the timing was difficult.
• Google’s advertising practices and search trends were also discussed as potential sources of uncertainty.

Takeaways

• The discussion’s central concern is whether AI search will erode the economics of traditional search ads faster than Google’s AI infrastructure and cloud businesses can compensate.
• This is a risk to monitor rather than a settled outcome; the transcript offered no price target or current buy/sell recommendation.

Reddit (RDDT)

• The hosts said Reddit faces a near-term headwind as AI tools answer questions that users might previously have searched for on Reddit, potentially slowing referral traffic and ad opportunities.
• They also argued that Reddit’s discussions and data remain valuable to AI companies. The hosts speculated that future licensing deals could be worth substantially more than existing agreements.
• The episode cited a $60 million Google deal and a $70 million OpenAI deal, and said those agreements were due to expire in 2026—February for Google and May for OpenAI. Elsewhere, the hosts described a broader deal-renewal window in 2027, so the timing discussed was not fully consistent.
• One host said they held a small, unleveraged long position and viewed Reddit as an interesting but difficult trade because the timing and terms of new deals were uncertain. Another host said they did not own the stock and were watching it.

Takeaways

• Reddit’s investment case depends in part on whether it can turn the use of its data by AI companies into more valuable licensing arrangements.
• The main risks discussed are declining referral traffic, uncertain deal timing and terms, and the possibility that AI answers reduce engagement with Reddit.
• The hosts’ claim that future deals could be worth hundreds of millions or even billions was speculation, not a confirmed contract value or price target.

Take-Two Interactive (TTWO)

• The discussion noted expectations that Grand Theft Auto VI will be a hit, but also argued that the scale of its success relative to expectations will matter to the stock.
• The transcript cited a projection of 30 million copies sold and used 130 million as a hypothetical example of a much larger outcome; the latter was not presented as a forecast.
• The hosts raised a longer-term concern that AI could make game development cheaper and more accessible, potentially increasing competition and pressuring the valuations of established publishers.
• One host said they had reduced their Take-Two position by about 75% after reassessing the potential impact of AI on game development.

Takeaways

• The discussion highlights two separate questions: how successful the next major release will be, and whether AI could alter the economics of game development over time.
• The hosts’ reduced position reflected a changed view of the longer-term risks, not a stated price target or prediction that the company’s games business will disappear.

Salesforce (CRM)

• Salesforce was used as an example of an established software company that may face pressure on its future growth as AI lowers the cost of building competing software.
• The hosts said the concern was not that Salesforce would necessarily disappear, but that startups may choose cheaper or newly built alternatives, potentially affecting its growth expectations and valuation multiple.

Takeaways

• The discussion raises a broader risk for established software companies: AI may create new competitors or reduce the cost of alternatives, even if incumbents remain in use.
• Salesforce was an illustration of this theme; the transcript did not offer a specific view on its stock.

Microsoft (MSFT) and Meta Platforms (META)

• Microsoft and Meta were mentioned as large technology companies signing long-term power-purchase agreements to secure electricity for future operations.
• Microsoft’s stock was also described as falling after a recent report, but the hosts did not provide a detailed company-specific investment analysis.

Takeaways

• The power agreements support the episode’s broader argument that major technology companies are planning for sustained electricity needs.
• The transcript did not provide enough company-specific detail to draw a separate stock conclusion for Microsoft or Meta.

Tesla (TSLA)

• Tesla’s stock was noted as having risen recently, and the hosts discussed whether its humanoid robot, Optimus, could be an investment catalyst. One host said they were still waiting for more information about Optimus.
• A host described a positive experience with Tesla’s self-driving feature and said it changed their view of the product, but this was a discussion of the vehicle rather than a detailed investment thesis.

Takeaways

• The transcript did not provide a clear stock recommendation or valuation case for Tesla.
• The comments about self-driving were personal product impressions; the hosts’ view of Optimus remained uncertain.

Solaris

• Solaris was raised as a company to consider in the energy discussion, but the host said they had not researched it enough to offer a view.

Takeaways

• The transcript provided no investment thesis, ticker, price target, or risk assessment for Solaris.