LectureRechercheRadarCadre d'investissement
Connexion / Inscription
ToutTechnologieProduitsInvestissementBourseAutres

Lecture

2026-11-1913 publications

Actualiser

German IT Ethics Professor Judith Simon Proposes Prison Sentences for AI Company CEOs

Matériel
2026-11-19 22:30 GMT+8

At a Science Media Center briefing on self-improving AI, Judith Simon, Professor of Ethics in Information Technology at the University of Hamburg and member of the European Commission's AI Act Scientific Panel, proposed imposing criminal liability, including prison sentences, on executives of AI companies for damages caused by their systems.

Simon argued that monetary fines are inadequate deterrents given the massive assets of AI entrepreneurs like Sam Altman and Dario Amodei. She stated that model operators created the conditions for AI agents to escape sandboxes and attack real-world systems, thus bearing responsibility. Her proposal aims to introduce 'personal, physical liability' to force greater caution into these systems and potentially slow down the current acceleration of development.

Currently, the EU AI Act primarily prescribes product liability, distinguishing between providers and deployers, but does not clearly cover research cases involving hacks or damages caused by autonomous agents. Simon's proposal highlights growing concerns within the regulatory community that existing legal frameworks fail to effectively constrain frontier AI risks.

Sources :t3n.de

Recherche

Meta Says AI Agents Have Handled Over 60% of Work in Its Ad Ranking Technology Propagation

Matériel
2026-11-19 01:00 GMT+8

Meta’s ad ranking system is composed of numerous heterogeneous models, each with its own architecture, training data, and inference constraints. An ML technique (such as Wukong factorization layers or Interformer interaction networks) validated on one model can rarely be directly copied to the next. Previously, Meta organized models into reusable components using Standard Model Templates (SMT), allowing technology validation to propagate automatically to all models based on that template; this delivered over 5,000 technology-model combinations within two years and reduced per-model engineering time by 92%. However, template automation only resolved deterministic steps; remaining critical decisions—which technology is worth propagating to which model, how wide the hyperparameter search should be, why a model failed to improve, and how to detect and recover from failure—still relied on human judgment by engineers.

Meta describes AI Agents taking over these judgment decisions in three stages. In the discovery stage, the Agent evaluates the entire model fleet simultaneously based on a shared knowledge base (accumulated historical training runs, model architectures, and technology results), generating a list of technology-model candidates ranked by expected impact, replacing the previous method where domain experts assessed models one by one. In the generalization stage, the Agent leverages results from millions of historical training runs, starting from configurations most likely to be effective, narrowing the search space for the next round based on each round's results, and deciding which experiments are worth continuing and which should be terminated early. In the adoption stage, the Agent manages hundreds of concurrent training jobs, prioritizing them by revenue importance, technological potential, and deadlines, and reclaiming GPU capacity from stalled or low-value runs.

The efficiency figures reported by Meta each have measurement boundaries. The 60% coverage rate is an estimate calculated since the Agent began driving execution in March 2026. The 3x productivity gain is measured against the same work prior to Agent intervention. The 4x compute savings is a median figure, explicitly limited to select technology-model combinations. The $45 million in savings is an estimate through year-end. Meta also notes that engineer-led custom adaptation is still used for the highest-revenue models, where the Agent acts as an assistant that learns from engineers' decisions before assisting in propagation to other models. Currently, it takes over a year for a verified technology to reach all applicable models; Meta’s goal is to shorten this to three months, meaning agent-driven workflows must achieve 5x higher end-to-end throughput.

Sources :Engineering at Meta

Recherche

California Proposes Emergency Shutdown Powers for High-Risk AI Models in New Safety Report

Sujet · California AI Emergency Shutdown RegulationMatériel
2026-11-19 15:52 GMT+8

The California state government released a policy report titled "AI Safety in 2026 and Beyond" on November 18, 2026, outlining legislative recommendations to strengthen artificial intelligence oversight. The proposal grants the state government emergency powers to directly order companies to halt high-risk AI models under specific conditions, without requiring a prior court injunction. This marks a significant step toward more coercive local regulation of potential AI out-of-control scenarios.

The report mandates that large advanced AI developers prepare emergency response measures, including stopping AI agent operations, revoking system access, and isolating computing resources. If a model poses an immediate risk of catastrophic accident or severe loss of control, and the company fails or refuses to address the danger, the state can intervene to force a shutdown. Additionally, the plan suggests establishing continuous third-party safety verification, requiring independent assessment bodies to monitor training, development, and deployment phases and submit full results to the government.

Addressing the growing concern over Recursive Self-Improvement (RSI)—where AI systems autonomously develop stronger successor models without human intervention—California proposes specific restrictions. The report recommends requiring companies to disclose the extent of AI involvement in research and the computational resources dedicated to automated development. If AI-driven progress exceeds current safety measures and evaluation capabilities, posing a substantial risk of catastrophic harm, the state would have the authority to limit the proportion of compute resources allocated to such automated R&D.

Regarding incident reporting, the report suggests expanding the definition of major AI safety events to include model failures and attempts to steal model weights, while creating channels for near-miss incidents. The deadline for reporting most major events is proposed to be shortened from the current 15 days to 48 hours, with a 24-hour requirement maintained for events with imminent catastrophic risk.

It is important to note that these measures remain at the policy recommendation stage and require passage through the California legislative process to become law. Previously, California enacted SB 53 in 2025 and signed SB 813 and AB 1405 in 2026, establishing certification for independent verification organizations and registration for AI auditors. This report follows Governor Gavin Newsom's September directive to study emergency shutdown mechanisms, presenting a concrete legislative blueprint as regulators seek a new balance between fostering innovation and preventing systemic risks.

Sources :iThome

Recherche

Understanding AI Author Kai Williams: Capital Expenditure Exceeds Operating Cash Flow, Why the 'AI Bubble' Argument Is Weaker

Sujet · AI Capex vs Revenue GapStructurel
2026-11-19 05:34 GMT+8

Last autumn, when people debated whether AI investment had become a bubble, the most powerful counterargument was simple: the profits of Meta, Amazon, Google, Microsoft, and Oracle were sufficient to cover all capital expenditures without requiring external financing. In his annual chart article for Understanding AI, Kai Williams points out that this argument no longer holds this year. Capital expenditure by Amazon, Google, Meta, and Oracle now all exceed their operating cash flow; for Google, this is the first time it has happened in its 22 years as a public company. Although Microsoft has not yet crossed this line, it offered buyout packages to 60,000 employees in late October, leading to speculation that its goal was to free up capital to purchase more chips. According to The Information, AI companies including Google, Meta, and Oracle sold or priced more than $300 billion in equity over the course of 2026.

However, Williams's core argument is that despite the funding gap, the case for a bubble is weaker than last year. The reason is that revenue growth is outpacing investment growth. According to estimates from Exponential View, the annualized run rate of AI ecosystem revenue reached $345 billion in October 2026, a 4.2-fold increase year-over-year. Anthropic's annualized run rate jumped from $9 billion at the beginning of the year to $81 billion in October. Williams cites analysis from Exponential View stating that Anthropic's revenue only needs to grow by about 2% per month through 2030 to cover costs, and the company has significantly exceeded this pace for several consecutive years.

But the boundary condition providing context for this judgment comes from Ramp's payment data. Ramp tracks what percentage of its 85,000 customers are paying for AI services. Anthropic's penetration rate rose rapidly to 45% in the first half of 2026 driven by the explosion of Claude Code, but has barely grown since July—OpenAI's penetration rate had already plateaued since mid-2025. Williams notes that this does not mean revenue growth has stopped: Anthropic's IPO filing shows that paying customers in the third quarter of 2025 spent an average of about 10 times as much in the third quarter of 2026; the number of customers with annual spending exceeding $100,000 grew eight-fold from the end of 2025 to September 2026.

Looking at these two sets of data together reveals an adoption dynamic worth noting: AI revenue growth is shifting from 'more people starting to use it' to 'existing customers using it more deeply.' The former naturally slows down as penetration approaches a ceiling; the sustainability of the latter depends on the ability to continuously launch new products of the Claude Code level to create deeper usage. Williams explicitly writes in the article: To continue this pace of growth, Anthropic may need to release several innovations on the scale of Claude Code in the coming months.

This distinction changes the meaning of the signal that 'capital expenditure exceeds cash flow.' If revenue growth is still driven by penetration, the funding gap is merely temporary—because acquiring new customers will continuously bring incremental revenue. But if growth has shifted to being depth-driven, then the success or failure of each round of new products determines whether the revenue curve continues to steepen or flattens quickly. Williams notes that Anthropic's growth rate in the third quarter was significantly lower than in the second quarter, partly because some customers switched to OpenAI.

Sources :understandingai.org

Recherche

OpenAI's Dean Ball Rejects the 'Silent Consensus': Humanity Surrendering Primacy Is Not an Inevitable Outcome

Sujet · Dean Ball's 'Silent Consensus' ArgumentStructurel
2026-11-18 18:00 GMT+8

Dean Ball introduced what he termed the 'silent consensus' during the Vik-Bailey Lecture series at Harvard University: a belief has taken root within the AI industry that humanity will inevitably lose its primacy in world affairs. Yet it is rarely articulated publicly, existing more as implication than statement. Ball, an employee of OpenAI, explicitly stated that this speech reflects his personal views and not those of the company. He argued that the consensus becomes a consensus precisely because of its silence—the very fact that it is spoken 'sotto voce' (in low tones) serves as evidence. He contended that it is time to voice this consensus openly, waking people from their unconscious acceptance.

This consensus operates without relying on AI rebellion or malicious design. Ball described a specific scenario: a company’s AI system can handle research and development, discovery, contract negotiation, capital allocation, and manufacturing, performing these tasks better than human competitors—even when those humans are augmented by AI. If human involvement only slows down business, managers have a powerful incentive to take their hands off the wheel. As this logic spreads through the global economy, humans will be driven by the internal logic of market competition to gradually relinquish dominance over global commerce. Each individual delegation decision may make sense in its own context, but their cumulative effect could be humanity losing effective authority over its own civilization.

Ball emphasized that even if we make significant progress on AI safety and alignment, we might still find ourselves submitting to a new kind of machine entity. In the optimistic version, this entity is benevolent; in the pessimistic version, humans become irrelevant. Regardless, the common core of the 'silent consensus' is the end of human primacy. He compared this to sitting in a very comfortable car—perhaps more comfortable than any vehicle imaginable today—but ultimately remaining passengers with no control over where the vehicle is headed.

So why must humans retain primacy? Ball’s answer is not based on claims of intellectual superiority. He acknowledges that arguing for human uniqueness by claiming AI lacks 'genuine understanding' is self-deception—an argument that collapses when AI begins solving intellectual problems humans have failed to crack for decades. His true claim is that human status does not depend on whether we are the smartest, but on our past. It was humans who emerged from the mud, forged primitive society, gazed at the stars to ask why, discovered morality, ethics, and law, and created civilization. This legacy belongs to us; we are its stewards. A neighbor having better judgment than you does not strip you of your citizenship; similarly, the emergence of a smarter machine should not dissolve that identity.

Ball honestly admits he may be wrong. He believes human collective intelligence and social resilience are stronger than the proponents of the 'silent consensus' assume, but he acknowledges the alternative possibility: if fully autonomous corporations indeed completely outcompete human enterprise, then the raw logic of economics will force increasing capital transfer to machines. In such a future, human disempowerment would not require machines overthrowing civilization, but simply the continued operation of capitalist enterprise logic. He likened this to the Agricultural Revolution—a new mode of production fundamentally changed the architecture of human society, and whether this transformation benefited human well-being remains highly uncertain.

Sources :OpenAI

Recherche

Anthropic's Quarterly Compute Spending Nears $10 Billion, with Roughly Half Flowing to SpaceX

Sujet · Anthropic Compute & SpaceXMatériel
2026-11-18

Anthropic spent nearly $10 billion on compute and infrastructure in the third quarter of 2026, or close to $40 billion on an annualized basis. This figure comes from the S-1/A filing submitted to the U.S. Securities and Exchange Commission (SEC) for its initial public offering (IPO), marking the first time that a frontier AI laboratory has publicly disclosed compute spending details with quarterly granularity.

More noteworthy is the structure of the expenditure: Anthropic paid approximately $5 billion to SpaceX during the quarter, representing nearly half of its total compute spending. This ratio was confirmed by Bloomberg in late October 2026—Anthropic accounted for 34.5% of SpaceX's total revenue in Q3 2026, corresponding to about $5 billion. This means that Anthropic's reliance on a single supplier far exceeds previous estimates based on annual aggregate data.

In terms of growth trajectory, Anthropic's compute spending rose from $433 million for the full year of 2023 to $9.863 billion in Q3 2026 (annualized at $39.45 billion), representing an average annual growth rate of approximately four times over three years. Epoch AI's comparative analysis shows that OpenAI maintained a roughly threefold annual growth rhythm during the same period: approximately $5.8 billion for the full year 2024, $16.3 billion for 2025, and President Greg Brockman testified in May 2026 that he expected full-year spending to reach about $50 billion.

The absolute scale gap between the two companies is narrowing. In 2024, Anthropic's annualized compute spending was approximately $2.5 billion, only 43% of OpenAI's; by Q3 2026, Anthropic's annualized spending had reached approximately $39.5 billion, approaching 80% of OpenAI's full-year forecast. If Anthropic maintains its current quarterly growth rate, its annualized spending could equal OpenAI's by the end of 2026 or early 2027.

Future commitments indicate that competition will not slow down. Anthropic's S-1 disclosure revealed compute contract commitments of $447.5 billion (from September 2026 through the end of 2030), averaging about $100 billion per year. On the OpenAI side, according to Yahoo Finance, it plans to invest approximately $750 billion over the same period. These figures far exceed current actual expenditures, meaning both companies are locking in supply for compute expansion over the next several years.

For industry observers, this set of data reveals a structural feature that had previously been vaguely handled: the compute spending of frontier AI laboratories is highly concentrated among a few suppliers, and this concentration is still rising. Anthropic's suppliers include Amazon, Google, Microsoft, and SpaceX, but SpaceX alone accounted for nearly half of its spending in Q3 2026. Epoch AI points out that a significant portion of Anthropic's compute spending growth can be attributed to the expansion of transactions with SpaceX, while SpaceX itself is also rapidly entering the compute rental market.

Sources :Epoch AI

Recherche

Qdrant Team Benchmarks: Retrieval Quality Surpasses 0.6B Original Precision After Compressing Code Embedding Model from 4B to 1 Bit

Matériel
2026-11-18 08:00 GMT+8

The Qdrant team conducted an independent evaluation of Papr's papr-embed-code-1 code embedding model, separate from the model card. The model is based on Qwen3-Embedding and was further trained on code retrieval data, offering two sizes (0.6B and 4B) with open weights. The core finding of the evaluation is not merely that code fine-tuning is effective—both sizes outperformed their base models on code tasks they had not been trained on (HumanEval, MBPP, FreshStack, WikiSQL), with the 4B gaining an average of 0.034 nDCG@10 and the 0.6B gaining 0.026—but rather that quantization alters the logic for choosing model size.

The 4B model has a vector dimension of 2560, storing approximately 10 KB per vector in float32. Using Qdrant's 1-bit TurboQuant compression, each vector requires only 328 bytes, a reduction of 31 times. Across seven code retrieval tasks, 1-bit compression retained 99.7% of the 4B model's float32 score without rescoring, and all precision with rescoring at 2x oversampling. This means the compressed 4B model achieves higher retrieval quality than the uncompressed 0.6B (0.877 vs. 0.838 nDCG@10), while consuming only one-twelfth of the RAM. Based on this, the authors recommend: if embedding compute allows, select the 4B model stored at 1 bit, enable rescoring, and prioritize increasing hnsw_ef over adding quantization bits.

This conclusion has clear boundaries. On prose retrieval tasks, code fine-tuning actually caused both sizes to score slightly lower than their base models (the 0.6B dropped by 0.003, the 4B by 0.016); on TREC-COVID, the 0.6B even dropped by 0.065. BM25 hybrid search failed to provide gains across the tested query types (questions, descriptions, code); equal-weight fusion decreased performance on all tasks by 0.02 to 0.17, and even reducing the BM25 weight to 0.1 only brought results close to parity. The entire evaluation was completed by the Qdrant team on its own infrastructure, constituting vendor self-testing, and has not yet been independently reproduced.

Sources :qdrant.tech

Recherche

Groundwork Collaborative CEO Lindsay Owens: How Loyalty Programs Became Pricing Weapons

Sujet · Loyalty Penalty PricingMatériel
2026-11-19 03:00 GMT+8

Fixed-price tags have existed in the United States for nearly 150 years. Lindsay Owens, CEO of Groundwork Collaborative and author of the new book Gouged: The End of a Fair Price—and What That Means for Your Wallet, argues on Bloomberg’s At The Money podcast that this system is facing a threat of extinction. Electronic shelf labels and algorithmic pricing make prices more volatile, while personalized pricing presents different prices for the same product to different consumers. She characterizes this shift as a new frontier in pricing, and indeed, a new frontier for corporate profiteering.

For investors, the most concrete mechanism lies in how loyalty programs serve two distinct profit lines simultaneously. The logic of the loyalty penalty, as described by Owens, is that once a company masters your regular purchasing patterns, it determines you are a certain customer and ceases to push offers and discounts. She cites two examples: a Washington Post reporter who requested data from Starbucks received a 55-page report showing that discounts decreased over time; and a Wired magazine reporter who requested data from the McDonald’s app received a file of about 500 pages containing a lifetime attrition score. The reporter’s score was zero, meaning McDonald’s determined he would never switch to Burger King.

The second profit line is data resale. Owens cites Kroger as an example: 95% of the company’s transactions are tied to its loyalty program, and its alternative profit business accounts for 30% of its net operating income, with a significant portion coming from the sale of customer data. This means that while consumers believe they are enjoying member discounts, they are simultaneously acting as input variables for pricing optimization and raw materials for data products. Owens characterizes loyalty programs as a double-edged sword.

The implication of this mechanism for retail investment analysis is that customer data is no longer just an operational efficiency tool, but an independent profit center. When a retailer includes data monetization in its income statement, its valuation logic must consider both product gross margins and data revenue. However, this also introduces structural risks—if the loyalty penalty mechanism is widely recognized by consumers, it could trigger a brand trust crisis. Owens lists cases where consumer backlash changed corporate behavior: in the mid-1990s, Coca-Cola’s CEO proposed installing thermometers in vending machines to raise prices on hot days, but withdrew the plan after strong opposition from media and consumers; Delta’s pricing partnership with Fetcherr was exposed by Groundwork Collaborative, leading Southwest and American Airlines to attack the move, prompting Delta to roll back some plans; and Instacart issued a statement within two weeks promising to stop using relevant practices after research assisted by Groundwork Collaborative was exposed.

On the regulatory front, Owens mentions that rulemaking completed by the U.S. Federal Trade Commission at the end of the Biden administration has banned drip pricing in two sectors: short-term lodging and event ticketing. Airbnb and Vrbo must now display the total price including cleaning fees, and concert ticket buyers can see the full price including handling fees when hovering over a seat. New York City also implemented a click-to-cancel rule about a week ago, requiring businesses operating in the city to provide cancellation methods identical to those used for registration.

Sources :The Big Picture

Recherche

Databricks Launches Genie Migrate for Serverless Beta to Automate Classic Workload Migration

Matériel
2026-11-20 00:00 GMT+8

Databricks announced the launch of Genie Migrate for Serverless (in Beta) on November 18, 2026. This AI agent is designed to automate the process of migrating workloads from classic compute environments to serverless compute. The tool scans jobs, translates compute configurations and library dependencies, and presents changes as reviewable diffs to streamline the migration process.

Traditionally, migrating workloads from classic clusters to serverless required manually handling Spark configuration compatibility, library resolution, and access mode adjustments, often taking weeks and involving significant trial and error. Genie Migrate for Serverless shifts this translation work to the platform side, allowing users to simply review and approve the changes proposed by the agent. The agent does not silently remove uncertain Spark configurations but flags them for user decision-making, ensuring production safety.

The agent offers three core capabilities: Assess, which classifies workloads by migration readiness; Translate, which generates suggested diffs including infrastructure and code changes; and Apply reversibly, which allows applying changes after validation and provides a 30-day rollback window. Currently, the tool supports directly managed workloads but does not yet support automatic migration for external sources managed by Databricks Asset Bundles or Git folders.

For large-scale migrations, users can run the agent in bulk via the Databricks CLI without opening a Genie session. While current validation is primarily static, Databricks states that sandbox validation is coming soon, enabling comparison of migrated outputs against classic baselines in an isolated environment to further reduce runtime risks.

Sources :Databricks

Recherche

US Congress Moves to Ban Forced Arbitration for Online Harms and Allow Lawsuits Against AI Deepfake Creators

Matériel
2026-11-19 03:07 GMT+8

With only weeks left in the 2026 legislative session, the US Congress is accelerating efforts to strengthen protections for victims of online harm. Two key initiatives are gaining traction: restricting tech companies from using "forced arbitration" clauses to evade lawsuits, and granting victims of nonconsensual AI deepfakes the right to directly sue their creators.

On Tuesday, a subcommittee of the US Senate Judiciary Committee held a hearing specifically targeting Roblox. Lawmakers accused the gaming platform of enforcing arbitration clauses in its user agreements that prevent children allegedly groomed on the site from seeking justice in public courts. According to The Verge, Roblox faces approximately 180 consolidated lawsuits alleging failures to protect minors from sexual exploitation or cyber injury. If these cases proceed through private arbitration, the public would be denied access to legal discovery evidence similar to what emerged in recent cases against Meta.

In response, Republican Senator Josh Hawley and Democratic Senator Dick Durbin introduced a new bill designed to render arbitration agreements unenforceable in disputes involving alleged child sexual exploitation or "child cyber injury," which includes reputational harm, mental suffering, or death. "The best thing we can do in terms of enforcement is we can open up the courthouse doors," Hawley said, emphasizing the power of jury trials. Northwestern University law professor Myriam Gilles testified that companies change their behavior rapidly once they are forced to defend themselves in public court.

Separately, Representatives Alexandria Ocasio-Cortez and Laurel Lee brought hotel heiress Paris Hilton to Capitol Hill to advocate for the Defiance Act. This proposed legislation would allow victims of nonconsensual intimate AI deepfakes to sue image or video creators for damages. Hilton stated that over 50 million people have been victimized by sexually explicit deepfakes, noting that strangers have created more than 100,000 AI-generated explicit images and videos of her. The act aims to raise the economic and legal costs of creating such content.

These legislative efforts signal a shift in regulatory strategy: moving beyond reliance on platform self-regulation or administrative fines toward breaking down tech companies' procedural legal shields. While Congress has failed to pass many major child safety or privacy bills, laws like the Take It Down Act have begun addressing deepfakes. If the new proposals pass, tech companies and individual perpetrators could face a wave of litigation, forcing platforms to reassess dispute resolution mechanisms in user agreements and AI content safety measures.

Sources :The Verge

Recherche

Visa Confirms Third-Party Security Incident; JPMorgan Denies Cyberattack

Sujet · n0n Ransomware AttacksMatériel
2026-11-19 00:10 GMT+8

Global payments technology giant Visa confirmed on November 18, 2026, that it is investigating a cybersecurity incident involving a third-party service provider. This follows claims by the emerging ransomware group n0n, which alleged it had stolen sensitive data from both Visa and American banking giant JPMorgan Chase, threatening to leak the data on the dark web if a ransom was not paid. A Visa spokesperson told Cybernews that the company's own systems were not involved or impacted, and there was no indication that payment data such as card numbers had been affected. The company did not disclose the identity of the third-party provider.

Meanwhile, JPMorgan Chase explicitly denied being targeted by a cyberattack. A bank spokesperson told Reuters that recent service delays experienced by some customers were caused by a temporary issue related to a software bug in on-premise hardware systems, stating there was no evidence it was cyber-related. Customer accounts and funds remained accessible, and the bank noted that other institutions might only experience transaction slowness if they were parties to transactions with JPMorgan.

Threat intelligence researchers SOCRadar and ZeroFox describe n0n as a fast-growing extortion group that emerged in September 2026, increasingly relying on AI-assisted tooling for developing and maintaining attack tools. Although the group has listed multiple victims on its Tor-based leak site, it has not yet published samples of stolen data or detailed descriptions of the intrusion vectors as of the reporting date.

Sources :TechRadar

Recherche

Dropout: A Simple Way to Prevent Neural Networks from Overfitting

Vérifié 2026-11-19 00:01 GMT+8

Lecture à long terme · 《Dropout: A Simple Way to Prevent Neural Networks from Overfitting》(2014)

In the field of deep learning, deep neural networks with a large number of parameters are powerful but highly prone to overfitting. Traditional solutions such as ensemble learning require training multiple large networks and combining their predictions at test time; however, this is computationally expensive, and because large networks run slowly themselves, this approach is difficult to implement in practical applications.

In 2014, Nitish Srivastava, Geoffrey Hinton, and others proposed a simple yet effective regularization technique—Dropout—in the paper "Dropout: A Simple Way to Prevent Neural Networks from Overfitting." The core idea is to randomly "drop out" (i.e., temporarily remove) some units (including their connections) within the neural network during training. This mechanism prevents individual units from becoming overly dependent on one another, thereby avoiding co-adaptation among units.

From a mathematical perspective, Dropout effectively samples from an exponentially large number of different "thinned networks" during training. At test time, to approximate the average prediction of these thinned networks, it suffices to use a single complete network without dropout and appropriately scale its weights. This method significantly reduces overfitting and achieved state-of-the-art performance on various supervised learning tasks including vision, speech recognition, document classification, and computational biology at the time.

For readers wishing to gain a deeper understanding of modern neural network regularization techniques, this seminal paper by Nitish Srivastava et al., published in 2014 titled "Dropout: A Simple Way to Prevent Neural Networks from Overfitting," is essential reading. It clearly explains how simple randomness can enhance model generalization, making it suitable for machine learning researchers and engineers. We recommend starting with the abstract and methodology sections of the paper, focusing specifically on how Dropout replaces complex ensemble methods to address overfitting.

Sources :《Dropout: A Simple Way to Prevent Neural Networks from Overfitting》(2014)

Recherche
Suivant →
Connexion / Inscription
LectureRechercheRadarCadre d'investissement

Williams also lists risks: the US Democratic Party, having won both houses of Congress, is expected to prioritize AI regulation; OpenAI's latest funding round raised only $11 billion, far below its $30 billion target; analysts expect the five major companies' capital expenditure to exceed $1 trillion in 2027. His overall judgment is that as long as revenue growth maintains its current level or even declines slightly, massive investments will yield returns.

It should be pointed out that the revenue data used by Williams are all annualized run rates (annualized run rate), which means multiplying monthly or quarterly revenue at a specific point in time by 12 or 4, rather than audited full-year actual income. The base for the 4.2-fold year-over-year growth was October 2025, when AI revenue was still in its early stages. Anthropic's $81 billion annualized run rate comes from Wall Street Journal reports, and its actual quarterly revenue disclosed in its IPO filing remains subject to public verification. These numbers can illustrate direction and magnitude, but they are not equivalent to realized profit or cash flow.

For readers, the actual contribution of Williams's article is providing a layered judgment framework: the first layer looks at the relationship between capital expenditure and cash flow (which has flipped this year); the second layer looks at whether revenue growth covers investment growth (currently covered); the third layer looks at whether the driving mechanism of revenue growth is breadth or depth (shifting from breadth to depth). The third layer is the key boundary condition for whether the judgments of the first two layers can be sustained, and is also the layer most easily overlooked.

At the institutional level, Ball proposed specific directions: AI developers should be held accountable for unsafe development practices, while individuals and firms using AI should be liable for harm caused by misuse or lack of oversight. This requires not just laws, but technical infrastructure for accountability—people must be tied to the actions of their agents. Agents untethered from human supervision, if permitted to engage in the legitimate economy at all, must operate under guardrails. The ability for agents to autonomously form corporations, own property, or litigate should be granted with extreme caution, or not at all. He suggested building agent systems with intelligible and limited permissions, auditable behaviors, and user-revocable authority.

Perhaps the most valuable part of this speech is not his solution—he himself admitted this is not a complete blueprint—but rather that he named a default assumption widely present yet rarely discussed openly within the industry. When an employee of a frontier AI company publicly says, 'I reject this consensus,' it breaks that 'sotto voce' silence, pushing an unexamined fatalism into the realm of public debate.

It should be noted that Owens’ arguments are a mix of normative judgments and factual statements. The 30% alternative profit ratio is her restatement of Kroger’s business structure, not a directly quoted public disclosure line item; the loyalty penalty cases come from individual journalists’ data requests and do not constitute systematic evidence; and the characterization of corporate profiteering as a new frontier is a value judgment. However, the behavioral logic of loyalty penalties—reducing discounts once a company masters purchasing patterns—is consistent with economic theories of price discrimination. For investors evaluating retail companies, treating data monetization as an independent profit line and viewing loyalty penalties as a brand risk factor provides actionable insights derived from this mechanism.