From July 24 to 25, three seemingly unrelated signals overlapped within the same time window: Anthropic's Claude was observed to show "thinking traces" shifting from a previously more complete display of reasoning to a form that was sparser or more condensed; SpaceX's Starship successfully completed its 13th test flight on July 24, Central Daylight Time, after three delays, with the first and second stages landing in the Gulf of Mexico and the Indian Ocean respectively while deploying 20 Starlink satellites; Donald Trump appeared in public wearing a cap that read "TRUMP 2028," claiming he had "won three times." Together, they point to a common thread: who controls the black box decides the boundaries of transparency. The change in Claude was first publicly noted by Professor Ethan Mollick of the University of Pennsylvania on July 25. He believes that even summarized thinking traces can help users diagnose model errors and provide insights, and that removing or weakening this layer of visual reasoning represents a significant loss of interpretability. As of now, Anthropic has not provided an official explanation regarding the reasons for the adjustments, the nature of its strategies, or future plans for explainability; in contrast, the Starship test flight maintained high-frequency public disclosures regarding key nodes such as mission delays, landing locations, and Starlink deployment, continuing SpaceX's long-promoted "test-fail-improve" iterative path, aiming to support lower-cost space transportation and Mars colonization visions with a fully reusable heavy launch system; Trump's assertion of "winning three times" does not refer to legally recognized election results but continues his narrative of a manipulated election, casting a shadow over the transparency of the electoral process through a vague implication of "TRUMP 2028." In this concentration of multi-dimensional events, the extent of the black box in technology and institutions has begun to be exposed to public view at a high frequency. How markets and voters "price" various levels of transparency and uncertainty is evolving into a core variable for technological risks, asset valuations, and policy expectations in the coming years.
Claude's Thinking Folded: Controversies Arise Over AI Interpretability Decline
On July 25, Professor Ethan Mollick of the University of Pennsylvania publicly pointed out that the "thinking traces" recently shown to users by Claude have undergone a noticeable change, shifting from relatively complete reasoning processes to a form that is sparser and more condensed. Currently, this observation remains centered on him as the core source. Mollick emphasized that even a summarized reasoning chain can help users understand the path through which a model reaches a conclusion, holding practical value when diagnosing errors in answers and discovering new independent insights. He views this adjustment as a "significant loss" for interpretability. In the practice of large language models, what is termed the Chain-of-Thought thinking trace has always been seen as a key means of enhancing the transparency of internal reasoning and the credibility of results, as well as an important signal for users to distinguish between "getting the correct results" and "truly understanding the issue."
Discussions surrounding the changes in Claude quickly focused on the trade-off between transparency and product form. Some discussions interpreted this adjustment as part of performance optimization or commercial strategy; however, as of now, Anthropic has not released any official technical explanations or policy statements regarding the reasons for the thinking trace adjustments, whether it is a long-term strategy, or future plans for explainability. In the absence of information disclosure, all external interpretations must remain at the level of assumptions and cannot be treated as facts. Currently, the topics of AI safety and alignment are gaining momentum, and how models are trained and constrained has directly entered the realm of regulation and public view. In this environment, technology companies must not only be accountable for output results but also bear a clear obligation to explain any reduction in interpretative channels and changes in interactive interfaces. How technology products maintain reasonable boundaries of transparency amid iterations is becoming an important observation metric for assessing their risk governance capabilities and long-term credibility.
Starship's 13th Test Flight: High-Risk Testing Iterating in the Open
Parallel to the discussions regarding Claude's thinking traces on July 25, another technological mainline presented in a completely different manner within the same time window. On the afternoon of July 24, Central Daylight Time, SpaceX's Starship completed its 13th successful test flight: the first stage landed in the Gulf of Mexico, the second stage landed in the Indian Ocean, and 20 Starlink satellites were deployed during the process, simultaneously validating payload, recovery paths, and commercial payload release capabilities in a single mission. Notably, this launch had previously been delayed three times, and the ongoing adjustments of the flight windows within a tight testing rhythm reflect the continuous trade-offs that complex engineering projects must make regarding weather, system status, and safety margins, rather than simple schedule delays.
More critically, SpaceX has repeatedly emphasized the "test-fail-improve" path concerning Starship—high-risk, high-cost flight tests are continuously exposed to public view, with landing locations, mission counts, and payload sizes iteratively accumulating in a traceable manner, providing a relatively clear risk-reward observation framework for the market and regulators. In contrast, the black box-style adjustment of thinking traces in Claude, captured by external researchers on July 25, lacks similar prior explanations and post-event data transparency: one exposes uncertainty to light, using verifiable mission results to stack long-term visions, while the other quietly tightens interpretative capabilities at the interface level. This difference in transparency is becoming a watershed for the long-term credibility of different technological paths.
"I Won Three Times": The Opacity and Emotional Mobilization of Election Narratives
During the same window from July 24 to 25, Trump chose to push "opacity" to the forefront in another way: he appeared in public wearing a cap emblazoned with "TRUMP 2028" and claimed, "I'm going to run again. It should be easy. I'm getting better at running for president. I won three times." This set of visual symbols and slogan-like phrases is widely viewed as a strong hint toward the 2028 presidential election, yet intentionally remains at the level of "hint"—as of now, there is no formally binding declaration of candidacy. "I won three times" mixes legally defined election results, the perception of "stolen victories" among his supporters, and the political myth he has constructed, continuing his previously emphasized narrative of "elections being manipulated." By continually questioning the transparency of vote counting, public disclosure, and other informational processes, he transforms the credibility of the system itself into a mobilization resource.
Against the backdrop of a prolonged American election political cycle, this narrative function does not even depend on specific election dates but normalizes "campaigning": by repeatedly releasing signals of re-engagement and victory narratives, it solidifies the emotional stickiness of core supporters and forcibly occupies the center of media agendas. From a data perspective, this forms a stark contrast with the technological logic underlying the Starship test flight and Claude's reasoning displays—the former is constituted by a verifiable chain of evidence made up of reproducible launch windows, landing coordinates, and satellite deployment numbers, while the latter adjusts thinking traces at the interface level, raising external doubts about interpretability. Trump's election discourse, on the other hand, is almost entirely detached from verifiable quantitative standards, relying on repetition, symbolism, and accusations of "black box elections" to shape his supporters' perceived reality. In this respect, the opacity of technological systems can foster disputes over risk assessments, while the opacity of political narratives directly rewrites the collective memory of the electoral process.
From Models to Rockets to Ballots: How Black Box Premium Reflects in Risk Pricing
Observing the three strands of Claude, Starship, and Trump within the same time window reveals that the technological and political black boxes jointly act on a core variable: how risk is priced. On July 25, Mollick pointed out the change in Claude's thinking traces from a single source, while Anthropic has yet to provide a technical explanation or roadmap, which causes the model's internal state to revert from "partially observable" to a state closer to a black box. For regulators and institutional investors, AI safety and alignment have become prominent issues; a decline in explainability means it is harder to provide clarity on accountability, compliance reviews, and reputation management, which typically is viewed as leading to potentially higher regulatory intensity and compliance costs, reflected in valuation models as higher discount rates or discounts applied to relevant business lines.
In contrast, on July 24, the Starship's 13th test flight was fully demonstrated to the public through livestreaming, landing locations, and results of the deployment of 20 Starlink satellites, with the risks of failure not concealed but rather incorporated into the publicly iterative narrative of "test-fail-improve." For capital markets, this high observability allows the technological progress of the heavy launch system, future launch costs, and business viability to enter specific models through parameters such as success rates in tests, transforming technological risk from "unknown risk" to "quantifiable risk," corresponding to a narrower interval for valuation distributions. Meanwhile, during the same period, Trump's narrative around the manipulated election was extended through his "TRUMP 2028" cap and claim of "I won three times," without providing verifiable institutional evidence; this kind of political opacity directly amplifies the uncertainties surrounding taxation, regulation, and industrial policies, often manifesting in asset pricing as higher policy risk premiums, transmitting into fields highly sensitive to institutional environments, such as technology, aerospace, and even cryptocurrency assets. The technological black box increases uncertainties at the compliance and operational levels, while the political black box amplifies uncertainties at the institutional and regulatory levels. Together, they make "black box premium" increasingly a cross-market pricing factor spanning models, rockets, and ballots.
The Black Box Will Not Disappear, but Transparency Is Being Priced by the Market
Looking back from July 25, 2026, at the signal overlap of July 24 to 25: Claude chose to reduce thinking traces for interpretability, the Starship's 13th test flight continued to demonstrate risk through public iteration, and Trump created a fuzzy space in election narratives with "I won three times" and "TRUMP 2028." The three strands of AI, aerospace, and elections present entirely different strategic combinations concerning transparency. For readers and investors, a workable assessment framework is: on the technological side, prioritize whether the model discloses its reasoning processes and error diagnosis capabilities; on the engineering project side, focus on whether testing delays, failures, and corrections are fully disclosed; on the political side, assess the degree to which narratives rewrite the boundaries of facts and the institutional uncertainties amplified thereby. Regardless of whether Anthropic reopens stronger thinking displays, whether Starship test flights maintain the same visible iterative path in the future, or whether Trump's hints ultimately transform into formal procedural actions, these choices surrounding transparency will feedback into prices through regulatory attitudes, capital flows, and risk premiums, forming a "black box discount" and "transparency premium" that are continuously priced by the market, while the ongoing struggle between technology companies and political actors over transparency, responsibility, and efficiency will long persist and continuously reshape this pricing structure.
Join our community, let's discuss and become stronger together!
Exclusive Hyperliquid benefits for AiCoin: https://app.hyperliquid.xyz/join/AICOIN88
Exclusive Aster benefits for AiCoin: https://www.asterdex.com/zh-CN/referral/9C50e2
On-chain Telegram community: https://t.me/AiCoinWhaleData
On-chain community: https://www.aicoin.com/link/chat?cid=N6OVMor5g
AiCoin on-chain Twitter: https://x.com/aicoinwhaledata
免责声明：本文章仅代表作者个人观点，不代表本平台的立场和观点。本文章仅供信息分享，不构成对任何人的任何投资建议。用户与作者之间的任何争议，与本平台无关。如网页中刊载的文章或图片涉及侵权，请提供相关的权利证明和身份证明发送邮件到support@aicoin.com，本平台相关工作人员将会进行核查。
