Summary

The key question around Yuanbao’s RMB1B red-envelope giveaway is not how it was allowed to launch, but why WeChat blocked it only on day 4, after traffic had already faded. 阑夕’s timeline: launch on Feb. 1, a suspected internal response circulating on Feb. 2 saying it was acceptable because there was “no threshold,” red-envelope decay on day 3, and a ban on Feb. 4. He invoked The Old Boys to ask: “It’s not why I locked you up for 20 years, but why I let you out after 20 years.”（不是我为什么要关你二十年，而是我为什么要在二十年之后把你放出来。） The core question is therefore how WeChat’s internal stance and boundaries changed, not simply whether it followed the rules.
Compared with WeChat’s 2015 red envelopes, Yuanbao bought short-term DAU but failed to create a sufficiently strong irreversible action. The key move then was card binding, not the size of the giveaway; Didi, Meituan, Pinduoduo and offline payment QR codes later embedded payments into everyday life. Yuanbao is now using sharing to unlock more draws, cumulative amounts and usage tasks, while the red envelopes quickly fell to a few fen. 潘乱 asked: “Isn’t this just something Pinduoduo played out years ago?”
Doubao, Qianwen and Yuanbao are using three different assets to compete for the AI entrance: Doubao is betting on experience and speed, Qianwen on Alibaba-wide orchestration, while Yuanbao has yet to answer what makes it indispensable as a standalone product. Doubao uses voice, multimodality and rapid feedback to cover older users, children and search scenarios; 潘乱 cited average daily consumption of 60T tokens. Qianwen is using RMB3B-scale investment plus Taobao Flash Purchase, Fliggy, Freshippo, medicine purchases and other services to train users into Agent-driven consumption. The show’s summary of Alibaba: “It’s not AI serving e-commerce; it’s e-commerce serving AI.”（不是AI要为电商服务，是电商要为AI服务。）
Yuanbao Pai embodies the one-sided idea of “if WeChat won’t arrive, build a WeChat ourselves,” but the current product looks more like a rushed experiment. 一泽 sees an attempt to mix real people, AI and communities, yet the product currently offers only 100-person groups, with no personal IDs, profiles or friend lists; even finding a previous private-chat contact is difficult. Retention and willingness to reopen the app after day one are also in question. The discussion compared it with the path dependence of the Yuanmeng Star era: “Yesterday once more”—features can be piled on, but no relationship or use case has yet emerged that makes people stay.
This year’s Spring Festival battle was judged by 庄明浩 to be unprecedentedly intense, with Tencent, Alibaba and ByteDance converging on the timing of the AI-entrance window. Doubao is sponsoring the Spring Festival Gala, Yuanbao is distributing RMB1B in red envelopes, and Qianwen has continued receiving resources barely two months after launch. 梁汝波’s call to “climb bravely to the summit” was linked to an AI wave that could be equivalent to “PC plus the internet,” but the show did not confirm any specific bidding or longer-dated market timetable.
Organization is determining whether AI investment can become a product: Alibaba can make the entire company queue up behind Qianwen, while Tencent must reconcile the fiefdoms and permissions separating WeChat, its models, Yuanbao and the various business groups. Qianwen’s model and app teams are also divided, but Alibaba is consolidating the base model and brand under the Chinese name “千问” and the English name “Qwen.” Yuanbao’s problem is not just budget; it is whether the product team can control model capabilities and whether the organization will sustain resource allocation.
The discussion’s three-month benchmark remains “Doubao far ahead, Qianwen catching up,” while Yuanbao’s ability to enter the race depends on the product, not the red envelopes. Qianwen’s upside depends on the model team, app team and business units forming a real coalition; Baidu is maintaining another funnel through its search front door, with the show citing roughly 700M MAU for Baidu App, 365M for AI Search and 200M for Wenxin Assistant. To B and To C remain separate operationally, but users may not divide model and brand perceptions along that line.
The longer-term investment question is not who spends the most during Spring Festival, but who can keep generating new products, talent and ways to fight. 一泽 said AI-era product managers must understand model boundaries and use a given scenario to produce a “step-change result”; 潘乱 criticized Tencent for being “not hungry for growth,” making it prone to turning a new campaign into a replay of the 2015 red-envelope war. The show’s closing judgment was: “Who wouldn’t want to return to 2015?” But going back would itself be “刻舟求剑”—mistaking an old solution for a new problem.

Deep dive

1. Yuanbao’s Red Envelopes Kick Off the Spring Festival AI War with the Oldest Growth Hack

阑夕’s first reaction was that the product felt “very old-school”: Chinese internet growth has become more restrained in recent years, while WeChat has long restricted incentivized sharing. A Tencent product spreading “so recklessly” through group chats and Moments was therefore highly unusual in itself.
潘乱 recalled seeing the mechanic on his way from an annual meeting in Guangzhou to his hotel and immediately concluded that it would flood everyone’s feeds: users received roughly 10 chances on entry, then 5 additional draws for each share. The number shown after the draw was only a cumulative balance, not cash that could be withdrawn directly. “Isn’t this just something Pinduoduo played out years ago?”
The low barrier and frictionless flow created rapid distribution, but the backlash was just as immediate. Large numbers of group owners proactively banned Yuanbao links, and 潘乱 warned people not to forward them in groups he managed. Many were already guessing on day one: “Long Ge sees this and shuts it down, right?”
一泽’s surprise came from the product, not the marketing. Yuanbao’s last major burst of attention came when it integrated DeepSeek in early 2025; in his view, its overall capabilities had not improved notably versus Doubao and Qianwen over the following year, yet it suddenly returned to the spotlight with an RMB1B giveaway.

2. Why the Day-4 Ban Matters More Than the Ban Itself

阑夕’s timeline was: red envelopes launched on Feb. 1; a suspected WeChat internal response appeared on Feb. 2, neither refuted nor formally confirmed, saying the campaign was acceptable because there was “no threshold”; red-envelope values began to decay on day 3; and the link was blocked on Feb. 4.
His question was not why WeChat banned it, but why WeChat waited until the peak interference had passed, after the number and value of red envelopes had clearly fallen. That points to an invisible variable: what internal negotiation took place over those four days, and how did 马化腾 view the episode?
The Old Boys analogy became the sharpest framework of the discussion: “It’s not why I locked you up for 20 years, but why I let you out after 20 years.” Yuanbao’s permission to spread is not the mystery; why that permission was withdrawn may be the real story.
“Long Ge may have returned from vacation” was merely the day’s popular joke. 阑夕 focused instead on the view that WeChat’s product gatekeeping remains in 张小龙’s hands. Even if Tencent’s senior management wanted to recreate 2015, it could not automatically change that product boundary.

3. “Equal Treatment” Holds, but “No Threshold” Looks Like a Slot Carved for One

庄明浩 cited WeChat PR head 周怡’s explanation: “User experience first, equal treatment,” accompanied by an image roughly meaning, “When I go crazy, I hit myself too.” 白鸦 emphasized WeChat’s long-standing operating discipline: familiar relationships can make communication smoother, but they cannot be used as a backdoor; WeChat also does not algorithmically monetize chat and Moments records.
阑夕 accepted the logic behind WeChat’s action but rejected the conspiracy theory that blocking Yuanbao was a way to conveniently shut down red envelopes from Qianwen, Wenxin Yiyan or the Douyin ecosystem. Even if other companies gave money away for free, they could not access WeChat Pay or wallet interfaces; users would have to enter their products to withdraw, creating a natural barrier.
The “no threshold” claim on Feb. 2 was therefore effectively a slot carved for one: only Yuanbao, or products used by Tencent’s senior management, were likely to have the conditions for direct payment inside WeChat. Whether this was unequal treatment mattered less than who controlled WeChat’s product rules and when those rules changed.
WeChat is also a “folding control”: post too much to Moments and it gets folded; post two comments in a row on an official account and they may also be folded. 阑夕’s summary was that whenever a type of information shows the capacity or tendency to become excessive, WeChat first handles it through product mechanisms.

4. Falling Red-Envelope Values Expose a Product Thinner Than Its Traffic

On day one, Yuanbao could still make users feel they were receiving RMB1 or RMB2; the amount quickly fell to a few fen. 潘乱 also questioned the necessity of designs such as cards with 100 codes. The distribution stimulus kept weakening, while the product actions left behind no comparable long-term asset.
潘乱 admitted that his judgment might be “easy to make from the sidelines,” but still felt Tencent had lost the pure product capability of the mobile-internet boom. Building a new product and getting it to a meaningful level of traction has become extremely difficult; often the only option is to have WeChat absorb the functionality into its unified ecosystem.
The later command-based red envelopes were especially damaging to Yuanbao’s product legitimacy in 阑夕’s view. 庄明浩 said the mechanic resembled polluting chat windows with garbled text; 阑夕 argued that Yuanbao had a certain special legitimacy from Feb. 1 to Feb. 3, but afterward looked more like the apps trying to freeload off Tencent’s ecosystem.
This growth campaign was not entirely devoid of usage tasks. 一泽’s review found that users needed to generate an AI-style image or ask a Spring Festival-related question to earn more chances to open red envelopes, but there was no red-envelope task directly leading users to Yuanbao Pai. He therefore leaned toward the view that the primary goal was still to build Yuanbao usage habits, not simply acquire users for Yuanbao Pai.

5. WeChat’s 2015 Red-Envelope Win Was About Card Binding, Not Holiday Buzz

潘乱 emphasized that the core metric in 2015 was an irreversible action: binding a bank card. Only then did the payment chain truly open; red envelopes were essentially transfers, and transfers accounted for a relatively small share of total mobile payments.
Didi rides, Meituan food delivery, Pinduoduo shopping and offline payment QR codes later put WeChat Pay into households and the “capillaries” of everyday life. Red envelopes opened the door; frequent, recurring use cases gave users a reason to stay.
Looking at today, Yuanbao’s RMB1B and Qianwen’s RMB3B-scale investment first address acquisition, DAU anxiety and user education. But the show did not see a permanent asset equivalent to card binding. Whether users return after the red envelopes stop is the question that the months after Spring Festival will have to answer.
庄明浩 also noted that when the 2015 red envelopes became a social event, regulators’ tolerance for that kind of social impact—or their understanding of it—may have differed from today. Even if Yuanbao genuinely replicated the scale of 2015, it could create additional risks the platform would rather not bear.

6. Yuanbao Pai’s Ambition: Grow Another WeChat Inside Tencent

一泽 extended the “card binding to payment scenarios” model: Yuanbao could first use RMB1B to bring people in, then let them discover groups, familiar or unfamiliar people, and a chat environment mixing humans with AI. Sustained interaction would then push data and retention higher.
His one-sided, deliberately extreme version was: “If WeChat won’t arrive, you can only build a WeChat yourself.” If WeChat continued to delay adding an AI entrance internally, Yuanbao Pai might try to become Tencent’s “third pole” after QQ and WeChat.
阑夕 rejected the idea that the Yuanbao team was waiting for WeChat to make a move. Its starting point, he argued, could not be waiting; it had to be running faster and better than WeChat.
一泽 also linked Yuanbao Pai to 马化腾’s “1B-person AI social experiment,” while qualifying that he recalled only “something roughly similar.” Tencent may want to observe on a new Discord-like platform which chat scenarios involving humans and AI catch on and which ones go cold.

7. Yuanbao Pai Has Piled on Features Without Growing Relationships

Yuanbao Pai groups can temporarily hold only 100 people, limiting the vision of channels or large-scale communities. The show further speculated that this could reflect a temporary retrofit or moderation and risk-control constraints. If the product was meant to host group-chat formats WeChat was unwilling to carry, it could theoretically test larger-scale chats, permissions and channel mechanics.
阑夕’s usage feedback was that Pai supports private chats, including private chats with everyone, but has no personal IDs, profiles, personal lists or friend system. If users cannot remember an avatar and name, finding someone they previously spoke with becomes difficult; relationship accumulation therefore conflicts with the product’s own feature set.
Family health reminders and reading partners are future possibilities, but 阑夕 is more focused on actual retention: does a Yuanbao Pai that users joined still have meaningful content after day one, and are they willing to open it again? That remains to be seen.
The Yuanmeng Star comparison was “Yesterday once more”: listening to music, watching video, chatting and building rooms are all theoretically possible, and the infrastructure is in place, but piling on features does not automatically create demand. Path dependence may instead conceal the absence of a core experience.

8. Tencent’s Premium Resource Allocation Has Not Reliably Produced New Entrances

庄明浩 listed four Tencent products that received premium internal treatment over the past six or seven years: Weishi, Call of Duty Online, Yuanmeng Star and Yuanbao. Weishi was once seen as Tencent’s lone answer to Douyin and received substantial resources. Call of Duty Online reportedly received Easter-egg ads triggered by WeChat chat keywords and perhaps some splash-screen inventory.
Yuanmeng Star was distributed at scale to compete with Eggy Party, after Tencent realized it had lost the entrance to younger users and the chance for many people to encounter games for the first time. But 庄明浩’s assessment was that the first three products were not particularly successful: Weishi largely failed, Call of Duty never broke out, and Yuanmeng Star generated decent traffic, while the real breakout was the later Delta Force.
Delta Force did not rely heavily on Tencent’s special resources and largely grew on its own. 庄明浩’s conclusion: “The iron has to be strong before it can be forged.” If the product is strong enough, ordinary promotion can amplify it.
潘乱 added that heavy promotion is still necessary. The question is which resources are used and whether they map to a product objective that can accumulate over time. The issue is not whether Yuanbao should be pushed, but what remains after the push works.

9. Qianwen Is Completing the Super-Entry Integration Taobao Never Managed

潘乱 said Qianwen had integrated a large number of Alibaba products in less than three months, including medicine purchases, movie tickets, Amap and the 扫街榜; Taobao, Fliggy, Flash Purchase, Freshippo, Taopiaopiao and Damai had also been placed inside its unified entrance. Many of these products had spent years without being fully integrated into Taobao.
一泽 argued that Qianwen’s RMB3B-scale investment was not simply about handing out money. It was about training users to order milk tea, order food delivery and make consumption decisions inside a unified chatbot entrance. Users would no longer open Taobao or Flash Purchase first; they would use voice to have an Agent complete the entire chain.
庄明浩 summarized the organizational direction: “What Taobao couldn’t get done, Qianwen will.” The 2023 narrative was Tongyi Qianwen integrating with Taobao to improve e-commerce efficiency. Today, Taobao, Fliggy and Flash Purchase are queuing up to enter Qianwen and serve the user.
This extends the mobile-era formulation: it was not wireless serving e-commerce, but e-commerce serving Alibaba’s wireless business. In the AI era, it becomes: “It’s not AI serving e-commerce; it’s e-commerce serving AI.” The change in subject determines how many resources Qianwen can orchestrate.

10. Alibaba’s Pyramid-Style Coordination Faces Tencent’s Fiefdoms

庄明浩 sees Qianwen as Alibaba’s most important product for the future: what Taobao cannot get done, Qianwen will. 潘乱 added that Alibaba initially assigned Quark to the task, but after becoming dissatisfied with Quark’s approach, started Qianwen separately, renamed the business unit Qianwen and continued investing—effectively giving the team a “sword from the emperor.”
阑夕 described Alibaba as a highly pyramid-shaped company, noting that 马化腾 had also expressed appreciation at an annual meeting for Alibaba’s ability to coordinate internal businesses. He further speculated that 马化腾 might privately envy that efficiency, while acknowledging that this was an inference about Tencent’s internal fiefdoms.
Tencent has long allowed business groups and lines of business to fight their own battles: WeChat operates as one system, games are divided among studios, and other businesses have their own boundaries. “Call headquarters when there is a problem; operate independently when there isn’t” once created vitality, but now makes it harder to build a cross-business AI entrance.
Qianwen itself has organizational divisions: the model team has its own showcase-style website, while the app is packaged and extended by another team; each business unit also has its own AI-transformation agenda. Alibaba’s move to unify the large-model brand under the Chinese name “千问” and English name “Qwen,” while retaining Tongyi Lab as the organizational name, is an attempt to resolve this kind of confusion.

11. Doubao Chose the Fastest Route to the Masses, Not the Longest Answer

Doubao’s product assumption is several hundred million DAU, so its answers are shorter and feedback faster; it does not ask older users or children to wait 30 seconds or even a minute. 庄明浩 compared it with early WeChat voice input and Scan: lowering the input barrier is how a product reaches a broader population.
一泽 compared voice input across Doubao, Qianwen and Yuanbao and found Doubao clearly superior. Its voice-input model has also improved by a step change over the past three or four months. According to people he spoke with internally, the prior setup may have had ASR and the LLM process inputs separately and correct one another, while Doubao integrated those capabilities into a single model that reasons and optimizes in real time during voice input.
The experience directly changes usage frequency. 一泽 placed Doubao on the iPhone side button, allowing him to hold it down and ask, “What is this?” or how to revive a dying flower. On mobile, speed is more likely than extended deep reasoning to become muscle memory.
庄明浩 also mentioned Doubao’s acquisition of the Douban database and integration of various POI datasets, covering clothing, fishing, taking children out, nearby restaurants and book recommendations. Doubao deliberately chose the fastest route to the largest audience rather than competing with Qianwen and Yuanbao on the length of its answers.

12. 60T Tokens Is Eye-Catching, but Audio and Video Offer a Shortcut

潘乱 added that Doubao’s average daily token consumption last year was 60T, a scale close to the Google Gemini web traffic token consumption he cited.
He also stressed that this cannot be equated directly with the number or efficiency of text users, because audio and video transmission consume far more tokens than text. Doubao is taking a different route through voice, images and video, so multimodal scenarios materially inflate consumption.
一泽 believes Doubao chose this path because it sees itself as a several-hundred-million-DAU product. Most people want rapid feedback, not to wait for a deep-reasoning model to produce the richest possible answer.
He linked this to the mobile-internet-era fight for time: Doubao and Volcano may now be competing not only for user minutes, but for a larger share of the tokens consumed across the user market.

13. AI-Native Entrances Will Shrink, While Back-End Capabilities Multiply

一泽 used Agent Skill to explain next-generation products: a Skill can teach a model to call APIs and use tools, or internalize an entire complex task and SOP. Users therefore no longer need to jump between structured forms, function pages and separate workflows.
Traditional AI note-taking apps typically ask users to enter text, upload images and then open a standalone chat window to call a pre-set Prompt. AI-native products let text, images and other multimodal materials enter through the same entrance, with results returned in the same streaming interface.
This is where Qianwen’s ability to order food delivery and purchase services matters: the front end is unified as far as possible, the back end connects a broad range of functions, and Agent-to-UI or the final output closes the loop. 一泽 summarized it as: “A unified entrance, a rich back end, and an end-to-end result returned at the finish.”
The show also mentioned Open Cloud providing a unified entrance on Feishu and Telegram, illustrating how users can submit different tasks in one chat interface and receive results after the back-end capabilities process them.

14. Tencent Already Has WeChat—Making the Case for Another Chatbot Harder

庄明浩 said major companies first have to answer whether to add AI to an old product or build a new AI-native one. The choice depends on each company’s resource base.
ByteDance and Alibaba did not originally have a dominant chat product, so they had to build Doubao and Qianwen from scratch. Quark had roughly 200M MAU but could not smoothly become a chatbot: old users still saw it as a browser and cloud drive, while new users did not care about that legacy. The company ultimately had to start a new product.
Tencent’s paradox is that it already has WeChat as a chat entrance. If the dominant AI product form is a chatbot, why not put the capability directly into WeChat, instead of having WeChat funnel users toward another product sitting on the same phone?
庄明浩 believes WeChat can “lie flat” on its social advantage and still recover users by opening selected AI capabilities later. But if it merely directs WeChat’s advantages toward Yuanbao, that is unlike the PC-to-mobile transition that captured a new entrance; it may instead amount to internal cannibalization. WeChat has introduced a Yuanbao bot labeled “AI” into chats, but opening the door to industry-specific AI contacts also creates security and moderation requirements.

15. The Thicker the Moat, the More Cautious the New Paradigm

潘乱 put Tencent and Apple side by side: both companies have such deep foundations that lying still is not immediately dangerous. A light combination of AI with existing businesses can already produce decent results, so neither faces the same need as Alibaba to seize a new entrance.
He noted that Apple spent relatively little on AI last year and shareholders were fairly satisfied, believing it had not wasted money. But Apple also posted the weakest gain among the “Magnificent Seven.” A thick balance sheet is a defensive advantage, but it may also reduce the appetite for deliberate self-disruption.
一泽 said Tencent’s strongest historical survival crisis came during the 360 episode: 360’s installed base could potentially uninstall QQ and Tencent software, directly threatening Tencent’s security around the entrance. Since the mobile-internet era, Tencent’s social and gaming businesses have remained relatively stable, and that sense of crisis has not persisted.
Alibaba, by contrast, must keep buying traffic and monetize through transactions and commissions. If a chatbot captures the user’s first attention and decision-making entrance, it could take share back from JD.com, Meituan and other platforms, giving Alibaba a stronger hunger for the entrance race. 潘乱 characterized ByteDance as focused on staying ahead, with AI central to its strategy from the beginning.

16. Foundation-Model Investment Has No Standard Answer; Baidu Gets a Free Ride on Search

庄明浩 asked which companies should still be making major investments in foundation-model R&D. Ant Group building 阿福 is easy to understand, but building 灵光 raises more questions. Meituan, Xiaomi and Xiaohongshu can explain investments in models, Agents, applications or verticals; there is no common answer for how far pure model R&D should go or how much it should cost.
潘乱 said Xiaomi’s model investment can at least be linked to cars and autonomous driving, with the possibility that the model eventually connects to a world model. As for Li Auto describing itself as an AI company and setting higher ambitions, he said he found that difficult to believe: a car company should first sell cars well.
一泽 offered another explanation: Chinese companies generally lack platform security. US companies can comfortably use a Google account to log into ChatGPT without worrying much that a competitor will cut off their account entrance. Chinese companies find it difficult to hand an intelligent cockpit or internal software entirely to another company, so self-development may serve as insurance against dependence on an external entrance.
庄明浩 cited QuestMobile data: Baidu App had roughly 700M MAU, Baidu AI Search roughly 365M, based on last August; Wenxin Assistant had another roughly 200M MAU, creating a “700M–365M–200M” funnel. Baidu’s advantage is that AI extends its core search business rather than being an extra AI feature added to another main product.

17. The Q2 Base Case: Doubao Leads, Qianwen Chases, and Yuanbao Waits for the Product to Prove Itself

Discussing the application landscape three months out, 一泽 predicted that “the overwhelming majority of people will choose Doubao,” while 庄明浩 summarized it as “Doubao may already be unstoppable.” Qianwen’s catch-up depends on retention, penetration and scenario frequency after Spring Festival, not merely on its red-envelope peak.
一泽 believes Qianwen must first solve the coordination problem across business units. Each department has its own AI-transformation agenda, while Qianwen is trying to provide a unified, more front-loaded entrance. The organizational structure will have to answer how the pieces are coordinated.
The show’s final shorthand was: Doubao competes on experience, Qianwen on marketing and the consumer ecosystem, while Yuanbao—the product that should theoretically be competing on product—has yet to find the right direction. 庄明浩 believes that after Yuanbao distributed red envelopes, its follow-up command-based red envelopes instead destroyed the product legitimacy it had previously built.
To B and To C remain distinct at the business level: enterprise services continue to operate as SaaS, and professional users differ from ordinary consumers. But users may not strictly separate “B-end models” from “C-end models” in their perception of products and brands.

18. AI-Era Product Capability, Growth Hunger and Generational Talent

一泽 believes AI-era product managers must understand model boundaries, define needs for specific scenarios and complete the solution. Doubao’s combination of voice input, decomposing vague questions, searching while thinking and multi-step search is a product decision, not simply a capability that emerged naturally from the model.
He also believes both large domestic platforms and startups need product managers who know how to “play AI extremely well”: use the model continuously in one scenario until it produces “a result that feels like a step change,” then package that capability into a vertical product. S Robotics Lab, Cloud Code and Cloud Code Work were cited as examples of probing model boundaries through product experiments.
庄明浩 worries that Tencent will struggle to tolerate a product growing slowly from a small base. He is a daily user of IMA and sees the team updating weekly and working hard to keep up with core categories, but still fears that model limitations could cause the product to fail at some point inside the organization. 一泽 believes IMA has performed well within its capabilities and that the main problem remains the model itself.
潘乱 criticized Tencent more directly for being “not hungry for growth”: it neither encourages buying traffic nor pursues enough new-product experiments, making it difficult for teams to develop new capabilities in real campaigns. ByteDance, by contrast, places more weight on data and the market. Products such as Tomato Novels and Hongguo Short Drama can continue as long as the market proves out, even if the founder does not personally use them.
The success of Delta Force has given Tencent’s old model positive feedback: small innovations built on mature categories can produce long-cycle products that last another 5 or even 10 years. Game outcomes are easy to measure and better at preserving team combativeness and hunger. Tencent places more weight on personal relationships; ByteDance keeps making room for younger people by moving older employees aside, with less weight on tenure and more competition.

(Preview) SaaSmageddon and the Future, Microsoft After a Market Correction, Anthropic’s Super Bowl Lies OpenClaw Debate: AI Personhood, Proof of AGI, and the ‘Rights’ Framework | EP #227