• Avoid trying to call a bottom in SMH or high-beta AI infrastructure stocks: the key risk is whether data centers can be financed and built, not just current chip demand.
• For NVDA, MU and AMD, wait for clearer evidence that demand and earnings are durable; monitor customer financing, data-center spending and memory pricing rather than treating recent declines as bargains.
• Favor diversified exposure over concentrated speculative positions; SPY or QQQ may be less volatile than individual AI-infrastructure stocks, though markets could fall further.
• UBER stands out as a cash-generative growth opportunity, with Amazon’s Prime delivery partnership adding reach; assess whether delivery and advertising growth can sustain momentum.

Semiconductor sector (SMH)

• The semiconductor ETF was down sharply while the broader market initially held up better, reflecting investors’ concerns about data-center construction, financing and the durability of AI-related growth.
• The discussion emphasized that strong earnings from chip companies were not enough to reassure investors; the market appeared focused on whether future growth can be sustained.

Takeaways

• The central question for semiconductor investments is whether planned data-center capacity can be built and financed—not just whether current demand is strong.
• The transcript cautions that a falling price or attractive valuation alone may not signal a bottom while sentiment and institutional risk appetite remain weak.

NVIDIA (NVDA)

• NVIDIA fell below $200 and later traded near $185, despite the host describing demand for AI compute as strong.
• The discussion raised the possibility that NVIDIA may need to help finance customers such as neocloud providers so they can afford its chips. The host and guest noted that this could prompt concerns about circular financing and put pressure on NVIDIA’s cash available for buybacks.
• The guest said he thought NVIDIA was less exposed to cyclical risk than Micron, but still viewed the data-center build-out and political response as important uncertainties.
• The guest argued that NVIDIA would have difficulty reaching a $10 trillion valuation soon if data-center investment were materially constrained. That was a conditional observation, not a price target.

Takeaways

• Monitor NVIDIA’s commentary on customer financing, data-center demand and the expected pace of capital spending.
• The discussion presents NVIDIA as a high-quality business facing uncertainty about how much of its expected growth can be delivered and funded.

Micron (MU)

• Micron fell below $700 during the sell-off, despite the host noting strong earnings and a low forward earnings multiple.
• The transcript described memory as cyclical and noted that some analysts expect earnings to peak in the coming years. That concern appeared to outweigh near-term supply constraints and pricing strength.
• Micron paused work on a $4.5 billion Taiwan plant amid a labor dispute. The host said reduced capacity could support memory prices, while also highlighting that construction and supply plans remain uncertain.
• A large purchase of Micron call options was mentioned, but the host treated it as an example of bullish positioning—not evidence that the stock had bottomed.

Takeaways

• The investment case depends on both memory pricing and how long elevated earnings can last; a low valuation based on peak earnings may be misleading.
• Watch for evidence that supply constraints translate into durable cash flow, rather than assuming reduced capacity automatically makes the stock attractive.

Advanced Micro Devices (AMD)

• AMD fell sharply after earnings, despite the host describing the results and demand outlook as strong.
• Its decline was discussed as part of a broader move away from semiconductor and AI-infrastructure exposure, rather than as a clear reaction to poor company results.

Takeaways

• The transcript suggests separating company execution from sector sentiment: strong results may not support a stock if investors are reducing exposure to the whole theme.
• Future data-center spending and customer financing are key issues to track.

Qualcomm (QCOM)

• Qualcomm fell about 11%–12% after earnings. The host said the company beat on revenue and earnings and had also announced a SpaceX-related deal.
• The discussion cited a weak smartphone market and uncertainty about the timing and scale of Qualcomm’s data-center opportunity.

Takeaways

• Qualcomm’s investment case appears to depend on whether data-center growth can become meaningful enough to offset pressure or uncertainty in its smartphone business.
• The transcript illustrates that an earnings beat may not be sufficient when investors question the durability of future growth.

Arm Holdings (ARM)

• Arm fell sharply after earnings, despite reported year-over-year growth in data-center royalties and the CEO’s comments about expanding its data-center CPU business.
• The host cited a high valuation—about 35 times sales and 138 times earnings—as a reason strong growth claims may not be enough to support the stock.

Takeaways

• Track whether Arm’s data-center business grows into its valuation; the discussion identifies valuation as a risk even if the business is gaining share.
• Smartphone demand remains another relevant consideration.

Broadcom (AVGO)

• Broadcom was caught in the semiconductor sell-off. The host said its valuation and outlook could also be pressured if AI capital-expenditure growth peaks earlier than expected.
• The discussion described Broadcom’s financing arrangements with OpenAI and Anthropic as important to their ability to buy custom AI chips.

Takeaways

• For Broadcom, assess not only demand for custom chips but also whether customers can fund large future commitments.
• The transcript frames an earlier peak in AI spending as a risk across the supply chain, not just for smaller semiconductor companies.

Taiwan Semiconductor Manufacturing Company (TSMC)

• TSMC fell with the broader semiconductor group, with the host noting a decline of around 5% early in the session and a larger loss later.
• The transcript did not identify a company-specific earnings problem; the move was presented as part of sector-wide de-risking.

Takeaways

• The discussion points to broad AI-capex sentiment as a near-term driver for TSMC, even without a specific negative company update.
• Watch whether data-center spending plans and chip demand remain intact as the political and financing picture develops.

Coherent (COHR)

• Coherent reported strong year-over-year growth in earnings and revenue, but its shares fell during the broader semiconductor sell-off.
• The host noted that good results were being discounted as investors questioned whether growth could continue.

Takeaways

• The transcript suggests that investors may focus more on the sustainability of growth than on a single strong quarter.
• Monitor future orders and demand tied to AI infrastructure rather than treating one earnings beat as a definitive signal.

Marvell Technology (MRVL)

• Marvell fell sharply amid wider semiconductor weakness and uncertainty around a possible acquisition of Lumentum.
• The host questioned the strategic and valuation implications of the potential deal, noting that competing bidders could add uncertainty.

Takeaways

• Consider both Marvell’s underlying AI-related business and the risks of a large acquisition when assessing the stock.
• The transcript identifies deal uncertainty and sector sentiment as near-term pressure points.

Lumentum (LITE)

• Lumentum fell from roughly $1,000 to near $500 in less than two weeks during the sell-off.
• The host discussed reports that there could be bidders for the company beyond Marvell, but said the possibility of multiple bidders added uncertainty.

Takeaways

• The discussion highlights how quickly a high-momentum infrastructure stock can reprice when deal terms and sector sentiment are uncertain.
• Any investment view would need to account for the acquisition situation as well as the underlying business.

Sandisk (SNDK)

• Sandisk fell sharply, at one point trading below $900 after having been much higher earlier in the year.
• The host connected its decline to investor concerns about the sustainability and cyclical nature of memory-related earnings.

Takeaways

• The transcript cautions against assuming a large decline makes a stock cheap when earnings expectations may also be changing.
• Follow memory pricing, supply and the durability of earnings to evaluate the outlook.

Dell Technologies (DELL) and Super Micro Computer (SMCI)

• Dell and Super Micro were cited as examples of server-related companies that had reported strong revenue but still faced sharp share-price declines.
• The host’s explanation was that the market was discounting a possible slowdown in the growth rate of data-center spending, even if current demand remained strong.

Takeaways

• For server and infrastructure suppliers, the trajectory of future orders may matter more to investors than recent revenue beats.
• The discussion suggests monitoring customers’ actual capital-spending plans rather than relying only on current sales.

CoreWeave (CRWV)

• CoreWeave was described as one of the most exposed companies to the cost and availability of financing. Its bonds were reported to be yielding around 17%, and the host said the debt market appeared increasingly skeptical.
• The transcript questioned whether NVIDIA or large financial firms would need to provide support to keep neocloud data-center expansion funded.
• Guest Steve said he bought more CoreWeave shares on the day, while acknowledging that the stock could fall further.

Takeaways

• The key issue raised is whether CoreWeave can secure financing at a cost its business can support.
• The guest’s purchase reflects his personal view, not a general recommendation; the discussion explicitly recognizes substantial financing and share-price risk.

Nebius (NBIS)

• Nebius fell below $130 and later traded near $115 as investors focused on the financing needs and construction plans of neocloud companies.
• The host said future earnings commentary would need to address the cost of capital, not only revenue, backlog or profitability.
• Guest Steve said he bought more CoreWeave, but did not say he bought Nebius.

Takeaways

• The transcript identifies access to affordable capital and the ability to build capacity as central factors for Nebius.
• Investors following the discussion would want to assess whether reported demand can be converted into funded, deliverable capacity.

IREN (IREN)

• IREN was identified as an upcoming earnings report that could offer a preview of investor sentiment toward data-center operators.
• The host said investors would be looking for discussion of capacity expansion and financing, and described the stock as sensitive to the broader data-center outlook.

Takeaways

• The transcript points to earnings guidance on financing, construction and capacity as important indicators—not just headline growth metrics.
• The discussion contains no specific price target or recommendation for IREN.

Applied Digital (APLD), Cipher Mining and TeraWulf

• Applied Digital, Cipher Mining and TeraWulf were mentioned among the smaller infrastructure names exposed to the AI and data-center financing environment.
• The host grouped such companies with more speculative names that could be vulnerable if capital becomes harder or more expensive to obtain.

Takeaways

• The discussion suggests greater financing risk for smaller or more speculative data-center operators than for established hyperscalers.
• Assess each company’s funding needs and ability to deliver planned capacity; the transcript does not provide company-specific earnings analysis for these names.

Alphabet (GOOGL/GOOG)

• Alphabet was among the large technology companies that initially rose as investors considered the possibility that slower capital spending could improve future free cash flow.
• The host cited Alphabet’s $15 billion data-center expansion in Malaysia as an example of companies potentially moving investment to more accommodating regions.
• Alphabet shares later fell with the wider market.

Takeaways

• The discussion presents Alphabet as relatively better positioned than highly leveraged infrastructure providers, while still exposed to regulation, data-center costs and market-wide volatility.
• Watch whether international investment can offset delays or restrictions in the United States.

Amazon (AMZN)

• Amazon was described as a large company that could benefit if a slower pace of investment improves free cash flow.
• The host highlighted Amazon replacing Grubhub with Uber Eats as its restaurant-delivery partner for U.S. Prime members, giving Uber access to Amazon’s large subscriber base.
• The transcript also mentioned an FTC allegation concerning Amazon’s advertising auctions, adding regulatory uncertainty.

Takeaways

• The discussion identifies both a potential free-cash-flow benefit from moderated spending and a separate growth opportunity in delivery and advertising.
• Regulatory scrutiny is a risk specifically raised in the transcript.

Microsoft (MSFT)

• Microsoft was among the hyperscalers that initially rose on the prospect that a slower spending trajectory could improve free cash flow.
• The host described Microsoft as having a more profitable cloud business and said free cash flow was expected to turn positive in 2027.
• The transcript also cited reports that Microsoft was tightening oversight of neocloud partners and had passed on some additional data-center capacity.

Takeaways

• The discussion presents Microsoft as comparatively well positioned, but emphasizes that investors should watch both its own spending plans and its reliance on external cloud partners.
• The 2027 free-cash-flow timing was discussed as an expectation, not a guarantee.

Meta Platforms (META)

• Meta was initially cited as a large technology company that could benefit from stronger future free cash flow if capital spending growth moderates.
• Meta’s president said building data centers would be harder and that projects require local support. She also said the company was active in 23 states.
• The transcript noted that Meta and other hyperscalers may consider building more outside the United States if domestic permitting becomes difficult.

Takeaways

• Monitor whether political and local permitting challenges delay Meta’s planned capacity or shift it overseas.
• The discussion views Meta’s scale as a potential advantage but does not treat its data-center plans as assured.

Apple (AAPL)

• Apple was described as a relative market leader during the rotation away from semiconductor and infrastructure stocks, though it later fell with the broader market.
• The host also noted concerns about rising memory costs, which could affect Apple and other large technology companies.

Takeaways

• The discussion portrays Apple as a perceived “safer” large-cap holding during a risk-off period, while noting potential exposure to higher component costs.
• Relative resilience during a sell-off does not, by itself, establish that the stock is immune to further declines.

Oracle (ORCL)

• Oracle was discussed as a legacy cloud company that may be better positioned than newer cloud operators with fewer established businesses.
• The transcript also raised questions about the capital required for Oracle’s data-center commitments and the potential impact of more expensive financing.
• Guest Steve said he was holding off on buying Oracle.

Takeaways

• The discussion suggests evaluating Oracle’s existing business base and funding needs separately from those of neocloud companies.
• The guest’s decision to wait reflects caution about the market environment, not a stated price target.

Uber Technologies (UBER)

• Uber reported strong earnings, and Amazon’s selection of Uber Eats as its U.S. Prime delivery partner was described as a significant distribution opportunity.
• The host highlighted Eats and advertising as businesses that diversify Uber beyond mobility and robotaxis.
• Guest Steve argued that Uber’s free-cash-flow generation made the stock more attractive than some speculative AI-related names.

Takeaways

• The discussion points to delivery, advertising and mobility as separate potential sources of Uber’s growth and cash flow.
• Investors may want to assess how much of the opportunity comes from the Amazon partnership and how sustainably Uber can grow outside robotaxis.

AppLovin (APP)

• AppLovin rose sharply after earnings that the host described as very strong.
• The stock had been heavily sold off beforehand, and the host cautioned that further work was needed to assess its advertising markets and growth prospects.

Takeaways

• The discussion shows that a strong earnings report can produce a sharp rebound after a major decline, but does not establish that the prior risks have disappeared.
• Further assessment of advertising demand and the company’s growth outlook was explicitly left open.

Figma (FIG)

• Figma fell after earnings despite the host saying it had beaten on several reported measures; the host was unsure what specifically disappointed investors.
• Its decline was discussed as an example of how market reaction may diverge from headline results.

Takeaways

• Review guidance and the details behind market expectations rather than relying only on reported beats.
• The transcript does not identify the specific concern driving Figma’s decline.

Celsius Holdings (CELH)

• Celsius reported earnings and revenue above expectations, with revenue up 17% year over year and earnings per share down 2% year over year.
• Its shares fell amid the broad market sell-off. The host did not identify a company-specific reason for the decline.

Takeaways

• The discussion suggests distinguishing company results from market-wide selling pressure.
• The transcript does not provide enough detail on Celsius’s guidance to draw a more specific outlook.

MercadoLibre (MELI)

• MercadoLibre reported a strong quarter but fell after margins came in weaker than investors wanted.
• The host described it as a long-term business with Mercado Pago as an important source of operating leverage, while saying margin improvement could take time.

Takeaways

• The discussion highlights the trade-off between growth investment and near-term margins.
• Monitor whether Mercado Pago and other operations eventually produce the operating leverage investors are expecting.

Grab Holdings (GRAB)

• Grab was described as relatively resilient during a broad market decline, with investors viewing its business as less directly tied to AI capital spending.
• The host noted that higher oil prices could still affect the company.

Takeaways

• The transcript frames Grab as a possible diversification from AI infrastructure exposure, but identifies fuel prices as a relevant risk.
• Its relative strength during one volatile session should not be treated as proof of lasting downside protection.

Palantir Technologies (PLTR)

• Palantir fell with other AI- and defense-related names despite what the host described as a strong quarter.
• The transcript cited concerns that political changes could affect data-center development and defense spending, while noting that the actual policy outcome was unknown.

Takeaways

• The discussion suggests that Palantir may be affected by political perceptions as well as its company results.
• Watch for actual policy and budget changes before treating political headlines as confirmed business impacts.

CrowdStrike (CRWD) and Palo Alto Networks (PANW)

• Both cybersecurity stocks were discussed as vulnerable to high valuations and to expectations that new AI models would create additional cybersecurity demand.
• The host suggested that slower AI-model development could weaken that narrative, even if the underlying cybersecurity businesses remain important.

Takeaways

• The transcript identifies valuation and the pace of new AI-model development as potential risks to the cybersecurity investment case.
• Separate demand for core security products from expectations tied specifically to AI-driven growth.

Bloom Energy (BE)

• Bloom was described as having reported a strong quarter, but its shares fell sharply during the sell-off.
• The transcript also noted Google’s planned Wyoming data center, which would use a large fuel-cell system from Bloom, though the news did not prevent the stock from falling.

Takeaways

• The discussion highlights that a customer project and strong reported results may not offset broad market de-risking.
• Follow whether announced projects translate into revenue and cash flow.

Tesla (TSLA)

• Tesla fell sharply during the sell-off. The host attributed the move partly to broader risk aversion and partly to the stock’s prior rise, which he said had been helped by optimism about SpaceX and robotaxis.
• The transcript also noted political concerns related to Elon Musk and the possibility of greater scrutiny by a Democratic-led Congress.

Takeaways

• The discussion identifies political sentiment, valuation and expectations around future growth as relevant risks.
• It does not provide a specific price target or a company-specific earnings catalyst for the decline.

SpaceX

• SpaceX was discussed in connection with financing plans for AI infrastructure and the possibility of a future public listing.
• The host noted that the company would need substantial capital to build data-center capacity and raised concern about possible share dilution and debt.
• Guest Steve said he was holding off on buying SpaceX and pointed to potential political scrutiny of Elon Musk.

Takeaways

• The transcript identifies funding, dilution, political attention and execution of data-center plans as key uncertainties.
• SpaceX is not a publicly traded stock in the discussion, and no ticker or public-market recommendation was given.

Rocket Lab (RKLB) and AST SpaceMobile (ASTS)

• Both were mentioned as high-beta space-related stocks that fell during the market decline.
• The host said Rocket Lab had dropped to its lowest level of the year at that point and noted that the stock had previously been much higher.

Takeaways

• The discussion presents these names as vulnerable to broad risk-off selling, despite their distinct business models.
• Consider the volatility and long-term execution risks rather than assuming past gains will quickly return.

IonQ (IONQ) and D-Wave Quantum (QBTS)

• Quantum-computing stocks were described as having fallen substantially during the year, despite remaining well above prices from earlier periods.
• The host emphasized that performance depends heavily on when an investor entered.

Takeaways

• The transcript illustrates the importance of entry price and volatility for speculative, early-stage themes.
• It provides no specific forecast or recommendation for either company.

Bitcoin (BTC)

• Bitcoin initially held up better than equities during part of the sell-off, then fell below $76,000.
• The host noted that crypto had not initially declined as much as some high-growth technology stocks.

Takeaways

• The discussion offers no specific crypto investment thesis or price target.
• Bitcoin’s brief relative resilience was not presented as reliable protection from a broader risk-off move.

S&P 500 (SPY) and Nasdaq-100 (QQQ)

• The S&P 500 and Nasdaq-100 fell sharply, with the Nasdaq-100 down around 6% at one point.
• Tom Lee was quoted forecasting a 15% rally in the Nasdaq-100 by year-end and an S&P 500 level of 8,200–8,400 by year-end. The host questioned how likely that outcome was after the sell-off.
• The transcript also discussed an average post-midterm rally of about 14%, typically beginning later in the year and extending into the following year.

Takeaways

• The transcript includes a bullish year-end forecast, but also underscores that it depends on sentiment improving and is not a certainty.
• For broad-market investors, the guest emphasized that index exposure can be less volatile than concentrated positions in speculative individual stocks.

Oil, gold and silver

• Oil rose about 1.5%–2% after reports of possible renewed U.S. military action involving Iran. The host said further escalation could increase uncertainty and pressure markets.
• Gold and silver were mentioned as rising modestly during parts of the sell-off.

Takeaways

• The discussion identifies geopolitical escalation and oil prices as potential sources of market and inflation risk.
• No commodity price targets or direct trading recommendations were given.

Berkshire Hathaway (BRK.B)

• Berkshire was cited as outperforming the broader market and holding a large cash position.
• The transcript described it as one of the relative areas of strength during the technology-led sell-off.

Takeaways

• The discussion presents Berkshire as a potential lower-volatility contrast to speculative growth stocks, but does not offer a specific valuation or recommendation.

Personal trades and portfolio-risk discussion

• Guest Steve said he bought more Ondas, QXO and CoreWeave, and was holding off on buying SpaceX and Oracle.
• He emphasized investing with money that is not needed for near-term expenses, avoiding leverage if unable to withstand volatility, and keeping some exposure to broad indexes or cash-flow-generative businesses.
• The host also stressed that many semiconductor and AI-related stocks had been under sustained selling pressure and that investors should not assume a rebound simply because prices had fallen.

Takeaways

• The practical lesson from the discussion is to match position size and investment horizon to the risk of the asset; the guest’s trades are personal decisions, not general recommendations.
• Key indicators raised throughout the episode were data-center permitting, construction timelines, customer financing costs, free cash flow and whether strong earnings translate into durable demand.