Vol.186: Kyth on Big Tech and Whether Podcasting Is Undervalued
Summary
Xiaoyuzhou did not begin by asking whether podcasts would become a mass-market product; it began with the unmet discovery and social needs of a highly engaged user group. The team noticed an active podcast community in 2018 and began interviewing users in August 2019; at the time, fewer than roughly 300-500 Chinese-language shows were updating consistently, while the US market and the listening scenarios enabled by wireless earbuds showed the category could expand. Kyth’s view: “What the market lacks more than anything is a role that aggregates scattered value and then amplifies it.”
Podcasts are difficult to scale exponentially, and Xiaoyuzhou accepted a long-term, roughly linear operating model from day one. One- to two-hour episodes are inherently difficult to share casually, and creators often need years of work before their efforts compound; instead of manufacturing spectacles around top stars, the platform uses editorial curation, timestamped comments, clickable Show Notes, and community interaction to create a “wasn’t this how it should have worked all along?” experience. For investors, that means a gentler growth curve and a platform more focused on long-term operations and community relationships.
Brands turn to podcasts for long-form communication, authenticity, and trust. As Kyth puts it, “Where else can you find a 60-minute brand communication environment?” On podcasts, that is routine; Xiaoyuzhou has tested platform content IP, brand co-creation, celebrity conversations, and brand-built podcasts. Users want both high-barrier information and emotional resonance, while senior corporate executives are overrepresented—helping explain why brands continue exploring the “ear economy” even under economic pressure.
The arrival of high-traffic platforms will raise the ceiling for podcast awareness and plays, but it will not automatically replicate Xiaoyuzhou’s community value. 李翔 worries that video platforms’ 100M-scale metrics could disadvantage a premium platform with “data fidelity” in front of creators and advertisers; Kyth acknowledges this must be solved, but argues that “video podcasts are still video,” competing for visual attention. Xiaoyuzhou’s more defensible advantages are audio’s relatively fairer expressive environment and the community experience of choosing where, and with whom, to listen.
Xiaoyuzhou puts operational health at least on par with growth, rather than copying the scale wars of food delivery and ride-hailing. 李翔 asks why first movers such as Ele.me, Yidao, and Dedao were often defeated by larger platforms; Kyth’s answer is that podcasting is not a purely utilitarian mass service, and a community cannot simply be swallowed by a larger player. The company’s immediate goal is to keep operating sustainably around break-even and let podcasting “gradually become something you can call a business.”
The stronger short video becomes, the weaker podcasts do not necessarily become, because people move compressed deep consumption into accompanying contexts. Kyth believes urban white-collar workers fit their desire to read, learn, and keep up with the world into a “media diet” consumed while commuting, driving, or cooking. Podcasts are not competing for the same screen; they are competing for accompanying time and providing an oasis in “a somewhat desertified era.”
AI will certainly lower the cost of topic selection, editing, video conversion, and content discovery, but for at least the next 5 years, human beings will remain podcasting’s scarcest asset. A US content organization already generates 3,000 AI podcast episodes a day, with the result described as “completely unlistenable”; AI can manufacture both sides of an argument at once, weakening the value of “I only want to hear your one, singular view.” Kyth therefore sees text-to-audio as a tool business, while the human limitations, vulnerability, relationships, and long-term companionship are the scarce qualities the platform needs to own.
China has yet to produce a Joe Rogan-level breakout host, and the primary variables are not traffic tactics but time, investment, and personality. Joe Rogan started podcasting in 2009 and had been doing it for 16 years by the time of the conversation; Kyth’s core advice to creators is first to clarify “who you are speaking to and what value they receive,” then embed that answer in the guests, structure, listening experience, title, and description. Podcasting may ultimately be more than an information product: it could be an “atmosphere machine” and a “trust machine”—a bridge built slowly through years of authentic expression.
Deep dive
1. Xiaoyuzhou’s first signal was not market size, but unusually intense user enthusiasm
Kyth recalls that starting in 2018, the team noticed content and discussion around “listening to podcasts together” steadily increasing within its existing community. What triggered the research was not a macro market report, but “a group of people investing enormous passion in one thing” while the supporting infrastructure was clearly inadequate.
From August 2019, the team spoke with podcast listeners and creators and distilled their needs into 2 words over 1-2 months: discovery and interaction. The first meant there was no effective way to find quality shows; the second meant that after listening for 1 hour, 1.5 hours, or even dozens of hours, users “didn’t want to close the App” and wanted to keep discussing the episode with creators and fellow listeners.
These 2 needs matched the team’s prior community-building capabilities and the profile of early podcast users. The decision process was not to prove the business model first, but to ask: “Why is your passion this intense?” Kyth believes that unusually strong but unmet enthusiasm may itself contain an opportunity.
2. Fewer than 500 consistently produced shows suggested supply could multiply many times over
Kyth estimates that in 2019, Chinese-language podcasts were mainly consumed on Apple Podcasts and Ximalaya, with fewer than roughly 300-500 shows being produced seriously and regularly, and even fewer creators having meaningful influence. The number cannot be precisely established from public data, but scarcity of supply was the team’s clearest observation at the time.
His line of reasoning was not “how many shows exist today?” but “where do these shows come from?” If a journalist or media professional could start one in their spare time, then the pool of people with comparable communication skills, professional expertise, and guest access was far larger than the existing creator base. Supply might multiply many times, leaving room for a platform role.
US podcasting was already much larger than the Chinese-language market in 2018. Apple launched AirPods around 2016-2017, and the noise cancellation of AirPods Pro later expanded listening scenarios. Wireless earbuds replacing wired earphones was a clear trend; technology did not create content directly, but it provided an important boost to the contexts in which content could be consumed.
3. The project was validated by users, not investors—and the team did not calculate the ceiling in advance
The company was then collectively searching for new projects, and interested members naturally gathered to discuss direction, differentiation, content, and community tone, while asking hosts and listeners in advance what the first version absolutely needed. Kyth says the phase was more about “letting passion drive the judgment”; the team did not rush to calculate the market ceiling or prove the business model.
李翔 asks whether they tested the idea with outside investors. Kyth’s answer is unequivocal: the earliest stage should be spent talking to users. Investors are not the best validators of embryonic demand; the team needed to find out what people far more passionate about podcasts than themselves actually wanted.
4. The first version had to feel like a mature player—and immediately feel different
As an RSS client, Xiaoyuzhou first filled in the basic capabilities of a podcast player, taking reference from third-party products such as Overcast and Castro, which Kyth had used for years. But playback was merely the ticket to entry; differentiated features determined the first impression.
The initial release supported timestamped comments, playback jumps directly from timestamps in Show Notes, and personal rankings by listening time. Kyth wanted a set of designs that would make users think, “Wasn’t this how it should have worked all along?” rather than features added for innovation’s sake.
The overall tone was built around being “clean, simple, and easy to understand.” Editorial picks were placed prominently because supply was still limited, quality density was high, and users had limited listening time. At a moment when algorithmic recommendations had become mainstream, human editors were better suited to express the idea that “the value of podcasts is vastly underestimated.”
5. Xiaoyuzhou chose to aggregate listeners first, rather than offer creators only a production tool
The team seriously compared 2 paths early on: building podcast tools for creators, or building a platform that helped creators gain traffic, distribution, and influence. Research led to a clear conclusion: “We had to choose the latter,” because a standalone production tool did not give creators enough reason to keep choosing it.
Kyth explains the mechanism this way: start with C-side consumption, aggregate listeners’ behavior, needs, and habits, then map user value into creator value. Only then can the 2-sided relationship turn over. Creators become willing to invest in their content, while listeners can interact with them productively in the same place.
This also determined the growth expectations. One- to two-hour shows are difficult to forward without friction, and a single show may need years before its work compounds. The team therefore never expected explosive growth; Xiaoyuzhou’s growth through its first 3 years, and even today, has remained broadly linear.
6. The 2020 remote launch turned every podcast into a new BBS
During the internal test on February 21, 2020, the team launched remotely from home, with everyone watching users arrive, data, and consumption behavior in collaboration software. Xiaoyuzhou went live on the App Store on March 26. Kyth says the backend signals remain “vivid” in his memory, including the first listener and the first imported show.
RSS solved content cold start. The real design problem was user cold start. The team contacted hosts across the ecosystem in advance and asked them to send invitation codes to their own listeners; users did not arrive through generic traffic, but followed shows they already trusted into a new listening environment.
The strongest validation came from the comment sections. Users turned their previously “full-to-burst feelings” into long comments and discovered that many others felt the same way. Kyth summarizes it as every podcast becoming a community, a BBS, with a new topic generated whenever the show updated.
7. Supply expansion came mainly from creators rolling in organically, not from the platform manufacturing star spectacles
In the early period and for several years afterward, Xiaoyuzhou rarely took non-podcast creators and actively remade them into podcasters. Media professionals saw peers doing it, realized they too had expressive ability, guest resources, and professional strengths, and entered naturally; supply accumulated through word of mouth, trends, and fit.
李翔 compares this with Douyin’s live-commerce model, which recruits top accounts and manufactures showcase events. Kyth admits the small team had neither those resources nor the belief that such a strategy suited the creation of a new category. “A forced match or a forcibly ripened fruit won’t be sweet.” The people with the strongest passion and desire to express themselves were the ones best suited to define podcasting’s early character alongside the platform.
8. Brands value 60-minute conversations and the trust accumulated through long-term subscriptions
Kyth believes 3 forces have converged in recent years: greater attention to Chinese-language podcasts, brands searching for new forms of brand advertising, and Xiaoyuzhou exploring monetization at the same time. A growing body of cases has gradually shown brands the platform’s user quality and community atmosphere.
Deep podcast content creates authenticity through sound, while subscriptions and long-term companionship build trust. The industry often asks, “Where can you find a 60-minute brand communication setting?” On podcasts, that is simply the normal marketing configuration; there is no need to compress everything into a few dozen seconds of copy.
As users increasingly need to sit with their emotions, podcasts also provide a setting for communication between people and between people and the world. Kyth has therefore become more convinced of the medium’s brand-building value, rather than seeing it merely as a quantifiable purchase of immediate traffic.
9. Monetization expanded from content IP to brands building their own “primary source”
This year, Xiaoyuzhou for the first time proactively designed platform IP around content: it discussed and deconstructed intimate relationships with Chanel for Valentine’s Day, and produced programming about family relationships for Mother’s Day. In these collaborations, the brand is not the protagonist, but appears as a supporter of good content.
More projects use brand co-creation. Tiffany’s Qixi collaboration brought a celebrity into a podcast, using a long conversation to show aspects unavailable outside the program. Other combinations included 杨天真 talking with 王嘉尔 and 蔡依林 launching personal podcasts, as well as 文淇 with 山下有松 and 贾玲 with Lululemon.
Kyth is especially receptive to brands producing their own podcasts. A brand podcast can become the brand’s “largest primary-source center,” fully unfolding its stories, ideas, and products. Bringing funding into the production is not inherently wrong; the question is whether the long-form content offers enough backstage and front-stage material to transcend ordinary promotional collateral.
10. High-quality information and emotional resonance together define the platform’s commercial audience
Kyth divides user demand into 2 categories. One seeks high-quality, high-barrier information, knowledge, perspective, and insight; the other seeks emotional value and spiritual resonance. Both require sustained listening and foster an ecosystem that is relatively tolerant of differing views and willing to engage constructively.
The platform observes a high proportion of senior executives across industries. 杨冰, CEO of Dewu, calls Xiaoyuzhou his “secret antenna,” while 李翔 calls it a “living room”—a place to invite different friends into deep conversations and keep discussing even when they disagree.
“Secret antenna” and “living room” are not demographic labels, but they accurately capture the consumption context brands care about. Users are not passively passing through an information feed; they actively bring shows into private time and give hosts sustained, undivided attention.
11. Xiaoyuzhou became strongly associated with podcasting through focus, not one-off marketing
Asked why so many people think of Xiaoyuzhou when they think of podcasts, Kyth resists over-attribution. He acknowledges only that the team’s understanding, care, and focus on podcasts may have played a role. Podcasting needs an “amplifier” that connects a content category to new listening and lifestyle patterns.
The fundamental difference of an independent product is that it serves a single audience. Podcast enthusiasts and podcast creators are the only people Xiaoyuzhou needs to serve. Listeners want the best possible service, while creators want to be easiest to find on a platform that keeps growing. The team therefore has to think more clearly and invest more completely than a generalist platform.
李翔 says Big Tech has 2 advantages: podcasting can be a “pet project” without revenue pressure, and it can access a huge traffic funnel. Kyth counters that truly return-free projects are rare, while flooding a product with traffic can damage its tone. As for complex functions, large platforms are fully capable of hiding them behind product design.
12. Top entrepreneurs appearing repeatedly on podcasts look more like industry milestones than creator count
Kyth is more interested in whether the most important deep content in vertical fields is increasingly being carried by podcasts than in the number of shows. Lei Jun appearing on 《高能量》 last year was an important milestone in his view. Since then, the heads of Nio, XPeng, and Li Auto have all appeared on podcasts, making long-form conversations with top entrepreneurs increasingly normal.
AI has seen a similar shift. Whether among the “Hangzhou Six Little Dragons” or other important founders, CEOs are not merely appearing on podcasts once; they are returning repeatedly. Kyth believes Xiaoyuzhou’s most viral single episode was probably 《严重话术》, in which 鲁豫 spoke with 张春. Emotional intensity, the chemistry of their expression, and the listening experience combined to create a powerful urge to share.
13. User expansion is inevitable; the key is to keep conflict within the right modes of encounter
李翔 points out that content communities almost all follow the same curve: as scale, creators, and commercial value rise, early users begin saying “the platform has changed” and may leave. He asks whether Xiaoyuzhou can accommodate newcomers while preserving high long-term affinity among existing users.
Kyth acknowledges that change is natural, but emphasizes that this is a question of degree, not whether change happens at all. Xiaoyuzhou’s comments and interactions occur under specific shows, making direct encounters between people with sharply different views less likely. Users first seek content; they are not thrown into a public square where everyone is forced into a head-on collision.
That does not mean the problem is solved. Kyth sees community governance as a content platform’s “job to do,” requiring a long-term combination of product, operations, and technology. Society may become more accepting of different kinds of people, but disagreement cannot be assumed to civilize itself.
14. Few creators quit outright, but busyness, low returns, and stalled growth are all real problems
Kyth observes that high-quality creators rarely stop updating completely. The 2 common reasons are being too busy and receiving less return than expected. The latter may refer to monetization, but it may also mean that after producing many shows seriously, neither category interest nor audience scale ever took shape, leaving the creator discouraged.
李翔 is skeptical of “too busy”: people capable of producing over the long term are unlikely to be genuinely unable to spare 1 hour a month. The problem is often that a show demands increasing investment and increasingly exposes the creator’s own life. Kyth agrees that sustaining verbal expression requires continuous updating, attention to listening quality, and ongoing exposure of personal experience; the creative burden does grow.
In the early podcast industry, especially during Xiaoyuzhou’s first year, creators were driven mainly by the desire to express themselves and by interest, with low expectations for financial returns. Among those entering now and in the future, the proportion treating podcasting as a content startup and money-making business will rise substantially. Kyth believes that after 1 serious year with no commercial return at all, it is entirely normal for most people to start reconsidering the project.
15. The platform must build income infrastructure and help new content reach uncovered audiences
The first layer of monetization is expanding total scale and influence. The second is improving C-side paid content, tipping, and later forms of direct payment, so hosts can earn returns after building listener trust. Revenue cannot depend solely on brand budgets.
On the B-side, the platform must keep educating the market that “podcasts are investable,” accumulate reusable cases and metrics for measuring campaign performance, and reduce the cost for brands and hosts to find one another. Kyth mentions platform initiatives including “追光,” which make it easier for creators to showcase their capabilities and for merchants to find suitable shows.
Listener growth cannot rely only on recommendation algorithms. The platform must identify content categories and audiences it has not yet covered, while creators need to develop differentiated angles within the same track. When a show creates content that did not previously exist, listeners may also follow it from outside the platform into Xiaoyuzhou.
In response to newcomers complaining daily that their plays are below expectations, Kyth offers no universal trick. Recommendation algorithms, content understanding, content containers, and discovery mechanisms are all still being rebuilt. The platform can reduce the friction of being discovered; creators still have to answer why listeners should choose them.
16. Big Tech’s arrival is both competition and free education for the podcast category
From other podcast clients emerging in 2020 and 2021 to Bilibili, Douyin, and Douyin Select encouraging long-form text and video, Kyth sees competition as an unavoidable stage of entrepreneurship. “If you work on something for a long time and discover that no one followed,” that may instead suggest the direction needs to be reexamined.
Awareness of podcasting remains low in the broader content market, so it is positive in itself for more platforms to make more people familiar with the term. Competition cannot be judged simply by saying “the other player is huge.” It requires breaking down what the new entrant is targeting, what it brings to the industry, which layer it is attacking, and how exactly it relates to Xiaoyuzhou.
李翔 agrees that large platforms can raise the ceiling. Several million plays is already top-tier on Xiaoyuzhou, while a video platform may easily reach a higher order of magnitude. But he worries that 100M-level topics on Weibo or video platforms’ statistical definitions could make a platform with “undiluted data” look less impressive to creators, consumers, and advertisers.
Kyth does not avoid the pressure around data presentation. He says only that this is a new problem that must be solved today. The platform cannot ask the market to ignore big numbers; it must turn real listening, relationship density, and commercial outcomes into value that people can understand and compare.
17. “Video podcasts” first obey video logic; audio may be fairer
Kyth’s first judgment is: “If you use the term video podcast, it is still video.” It may be created as a podcast-style conversation, but ultimately it competes with other videos for users’ eyes and is governed by video platforms’ distribution, viewing, and commercial logic.
The video wave will still bring creators and listeners to podcasting, and some will stay. Others will discover that continuous filming is too burdensome, that facing a camera makes them less articulate, or that their comfort zone simply is not visual. They will eventually return to the audio format that better suits their expression.
Compared with video, audio has smaller person-to-person gaps in production cost, polish, and absolute attention-grabbing power, and may therefore be a “fairer creative environment.” Kyth keeps the conclusion open: creators will test whether to make every episode a video or only occasional episodes, whether clips can drive distribution and sponsorships, and Xiaoyuzhou may still take action around video in the future.
18. Food-delivery-style scale wars are not a complete analogy; community and operating health change the conditions of victory
李翔 cites 3 historical pressures on first movers: Meituan raised the ceiling for food delivery and Ele.me lost; Didi scaled ride-hailing and premium service provider Yidao was defeated without receiving a strong return; Dedao was once asked how it would respond to Ximalaya’s broader content and larger user base. Being a pioneer, preserving a premium tone, and ultimately winning are not naturally equivalent.
Kyth’s breakdown is that food delivery and ride-hailing eventually became absolute mass markets, with highly utilitarian core value and almost no community or relationship between user groups. Xiaoyuzhou has content and community characteristics; users care not only which app they use to listen, but also “what kind of people they are listening with.”
He is careful not to call community a moat outright, but believes it makes users care about the environment they are in and the people they are consuming alongside. That creates an opportunity for content platforms with community characteristics, though it does not mean a large platform can never absorb a smaller one. For Xiaoyuzhou today, it is better to run the company well and operate sustainably around break-even than to fight over absolute scale; health matters at least as much as growth.
李翔 adds that Xiaohongshu shows a platform with a distinctive tone can still grow while facing larger flows of traffic and capital. Kyth agrees that this is a successful outcome, but does not turn the individual case into an inevitable path.
19. US video podcasts provide information, but cannot serve as a direct road map for China
US video podcasting is primarily carried by YouTube, whose strength in traffic and monetization is close to monopolistic. Xiaoyuzhou has neither the same platform conditions nor grounds to assume that Chinese-language podcasting will be “nourished” by an equivalent player.
Kyth also notes that the growth of leading US hosts is connected to local political agendas, topics the public cares about, and a shift in emotional and expressive formats away from traditional television shows toward podcasts. Which of these conditions can be replicated in China, and which do not exist, must be examined one by one.
He therefore does not see video as a shortcut that would make Xiaoyuzhou “become another company overnight.” The more realistic question is how to extend Xiaoyuzhou’s existing value under the conditions of the Chinese internet, its existing audiences, and its community tone, building a healthy and self-consistent ecosystem rather than copying YouTube’s answer.
20. Short video captured the eyes and pushed deep content into the ears
Kyth believes that the more deeply short video reshapes life and content habits, the more likely the same person is to develop an “endorphin need”: after realizing they have not read a book or a long article for a long time, they try to replenish their “media diet” and perform a kind of self-rescue.
As waking hours are increasingly contested by high-stimulation content, urban white-collar workers move keeping up with the world, expanding their horizons, and self-improvement into accompanying contexts such as driving, taking the subway, and cooking. Xiaoyuzhou grew precisely as short video and short-form content became stronger; the 2 formats are not competing for exactly the same time.
Kyth likes to say that podcasts provide an oasis in “a somewhat desertified era.” They occupy accompanying time, but also let users feel again that communication between people can be whole and comfortable, rather than continually cut into shorter, more stimulating fragments.
21. 10x growth requires more needs to become podcast-native, not simply more copies of existing shows
If the user base were to grow by another order of magnitude, Kyth believes the platform would need to provide more practical or entertainment value and serve people whose needs are not currently met by podcasts. Demand-side change is ultimately led by supply: activities currently completed elsewhere first need new supply suited to audio.
“Learning becoming podcast-native” is his example. News and industry shows already perform part of the learning function. In the future, university knowledge may not have to be delivered as courses; scholars or students could publish weekly, using a gentler, more relaxed, lower-burden format to provide ongoing companionship and address broader learning needs.
He does not seek to define podcasting with extreme precision, retaining only a few essentials: it is centered on speech and verbal expression, can involve 1, 2, or more people, gives listeners the value they need, and updates continuously. The medium, channel, and content will generalize, but sustained personal expression remains central.
22. A breakout host compounds over 16 years; the creative method is a spectrum from interviewing to conversation
Asked why China still lacks a universally recognized Joe Rogan-style host, Kyth’s first answer is that “there has not been enough time.” Joe Rogan started in 2009 and had been doing it for 16 years by the time of the conversation. His consistency, investment, mass appeal, and talent for bringing any guest down to angles ordinary people care about allowed his personal brand to compound like a snowball and pull far ahead of the second and third players.
Lex Fridman’s advantage is closer to a path the Chinese ecosystem can catch: mastering a guest’s professional field deeply enough to conduct a genuine conversation. 李翔 adds that a lead of this magnitude may also be the result of individual genius, like Márquez and One Hundred Years of Solitude, and cannot be attributed entirely to industry maturity.
For newcomers, Kyth retains 1 overarching recommendation: first clarify “who you are mainly speaking to and what value that audience gets from the show.” The answer should run through the content structure, listening experience, guests or co-hosts, title, and description, rather than relying on distribution to compensate for vague positioning after launch.
In terms of format, he believes audio’s low cost allows more monologues and relaxed, casual experiments that feel less like finished “works.” The platform prefers to “discover and amplify” rather than overstep by designing a show from zero to one. Audio dating shows and audio livestreams with virtual avatars both suggest that many formats remain unexplored.
23. Podcasting is not a fixed interview technique, but the long-term production of relationships, vulnerability, and atmosphere
Kyth’s advice for 《高能量》 is for 李翔 to add more personal, subjective, and even vulnerable expression. The voice of a mature host can easily acquire a slightly downward, condescending quality; sincere remarks that would not fit a grand formal setting are often more memorable and make listeners feel they have seen the whole person.
李翔’s reservation is worth preserving: society should first evaluate someone through the responsibilities of their role, just as a doctor is judged by ethics and medical skill, and an entrepreneur by how the company treats its employees and customers. A person should not receive extra credit simply because they cook well or have an interesting private life. As a journalist or interviewer, he cares more about whether he has fulfilled his professional role.
Kyth responds that a podcaster’s role need not be identical to a journalist’s. He used to say that “a podcast is not an interview,” but has changed his view: interviewing and conversing form a spectrum. A host and guest can operate at 1:9, 2:8, or 3:7, with no universally optimal ratio. Familiar recurring partners are more likely to be consumed as a relationship, while unfamiliar guests may be used primarily for information.
李翔 distinguishes CCTV’s Dialogue from Qiang Qiang San Ren Xing as formats led by interviewing and conversation, respectively. He uses Robert Rubin’s style of chairing meetings—asking almost only the key questions—to show that restraint can also be a form of role fulfillment. Kyth compares a multi-person podcast to a disco: every mouth is an instrument, the host is the DJ, and listeners want the hour to feel comfortable and satisfying enough that they are willing to return to the room.
24. As AI manufactures more content, human limitations become a scarce asset for the years ahead
Kyth believes AI will “undoubtedly” improve efficiency in topic selection, production, editing, cover art, audio-to-video conversion, speech-to-text, summarization, and content discovery. But automatically turning a report into a Q&A audio program is closer to an audio-processing tool than to the core value of the podcast ecosystem.
A US content organization already uses AI to generate 3,000 episodes a day in an attempt to arbitrage platform ad revenue. Industry feedback, however, is that the output is “completely unlistenable.” Text farms can be skipped at a glance; low-quality audio consumes linear time. If both the conversation and script are generated by AI, the creator’s value is almost entirely replaced.
AI can manufacture infinite arguments for and against, but listeners with limited time would rather give it to “a real, singular point of view.” Kyth expects that for at least the next 5 years, the atmosphere, emotional exchange, comfort, and ability to move people in human conversation will remain difficult to replace. Listeners will choose for themselves between human and AI shows, and platforms will govern them accordingly.
In 10 years, podcasting may become the “personality card” of anyone who wants to express themselves; after listening for 3-5 hours, it may feel as though you have already met the person. Learning, socializing, film and television, sports, and consumption may all become podcast-native, while production costs continue to fall. What remains truly valuable is “authenticity and connection”: listening to someone speak the truth in their head over the long term, allowing trust to accumulate like a bridge, and ultimately turning podcasting into a “trust machine.”