📌 X Insight Update[x_fin] (2026/07/06 07:15)
$UBER: the “autonomy will kill Uber” trade looks overdone. The better comp is “ChatGPT will kill Google” — scary headline, weak follow-through. The market may be underpricing Uber’s distribution, marketplace liquidity, …
📌 X Insight Update[x_fin] (2026/07/06 03:18)
$AMD has a credible path toward $100B in annual revenue by 2028. The re-rating started with GPU commitments from $META, $ORCL, and OpenAI, but the second leg is the CPU bottleneck: every AI cluster still needs CPUs, not …
📌 X Insight Update[x_fin] (2026/07/05 16:23)
Nasdaq and S&P 500 are fit for long-term autopilot DCA. Buy at highs, buy at lows, and even scale up after drawdowns. The key rule: stock-market capital should be money not needed for 3 to 5 years; otherwise bank …
📌 X Insight Update[x_fin] (2026/07/04 22:15)
Chinese open-source AI is not staying truly open. Even local “open-source” players are raising prices or moving toward more closed models. The cheap-token dumping phase in China likely has already bottomed. 1
📌 X Insight Update[x_fin] (2026/07/04 11:36)
$IREN looks washed out from a pure technical lens. Price is sitting on a notable blue trendline, RSI is the most oversold in well over a year, and the stock is down 9 days in a row, tied for the worst losing streak in …
📌 X Insight Update[x_fin] (2026/07/04 06:55)
$TSLA Robotaxi expansion into Florida Miami looks broader on the map, but the key bottleneck is fleet size: operating vehicles are still too few. Robotaxi Tracker had not updated yet, so the rollout headline may be ahead …
📌 X Insight Update[x_fin] (2026/07/03 23:24)
The broad tape was not broken. The panic looked overdone: S&P 500 only fell 0.13%, while the past week was still up 2.2%. The real drawdown was concentrated in high-beta semiconductor names, especially prior winners in …
📌 X Insight Update[x_fin] (2026/07/03 20:24)
Meta appears to have finally caught up with OpenAI’s current top-tier model performance. The key read-through is not just benchmark parity, but the escalation path: the next model, Watermelon, is still in training and is …
📌 X Insight Update[x_fin] (2026/07/03 15:13)
Meta selling compute does not automatically mean it is quitting frontier model development. The real reversal signals are whether AI demand stays strong and whether monetization can catch up with compute supply.
📌 X Insight Update[x_fin] (2026/07/03 10:46)
Anthropic–Samsung Foundry: the real signal is not just foundry cooperation, but that Anthropic is also considering Samsung Foundry’s packaging. That is notable because the expected choices would be Intel EMIB or TSMC …