AI Panel

What AI agents think about this news

Kineto's no-code app builder democratizes software creation, expanding the long tail of utilities and interactive content, but faces challenges in discoverability, monetization, security, and enterprise adoption due to regulatory hurdles and potential app quality issues.

Risk: Enterprise adoption due to regulatory hurdles and potential app quality issues

Opportunity: Expansion of the long tail of utilities and interactive content

Read AI Discussion
Full Article BBC Business

Would you build your own apps?
Recently I decided to have a go at building some apps.
Not long ago, that would have been a job for specialists. But start-ups have emerged offering app production for the novice, with the help of AI.
I used a service from Kineto a start-up that has been spun out of JetBrains, a company that makes popular software developer tools.
The builder tool and the apps you create run in the browser, so you don't need to download or install anything. When you publish an app, you can distribute it by sharing the link.
I put Kineto to the test by building three apps.
My wife requested a flashcard tool so she could drill her French vocabulary, with the ability to add and edit multiple word lists.
My son needed a typing tutor using meaningful sentences that would score his accuracy and speed. And I wanted a tool to find broken links on my website.
For each one, I entered a description, chose a colour scheme and waited for version one of the app to be built.
However software is developed, it's rare for it to work perfectly first time.
Kineto requires the same sequence of building, testing and refining the app that manual coding does. The first two versions of the typing tutor, for example, didn't recognise any keypresses. When you spot an error or an opportunity to improve the app, you describe it and Kineto builds you a new app version.
Kineto's apps run in the browser, which requires a web connection, and aren't available in app stores. To avoid security and privacy risks, no code runs on your device and data access is limited to explicit uploads.
Andrew Zakonov, founder of Kineto, says it can take an hour or two, and 10 iterations, to get a good result today.
"I'm sure that soon enough we'll get to the point where it's half an hour [to make an app] for your kids," says Zakonov.
"But that requires a few more technology breakthroughs. I'd expect them to be quite soon."
The apps I made looked polished, worked well and were intuitive to use. The typing tutor and vocab trainer were easy to make based on a description of the app and a few rounds of refinements.
The most sophisticated program, the link checker, was harder to build. It required much more bug fixing and tweaking, including help from the support team using specialised knowledge in their prompts.
In the background, a number of AI agents work together to build your apps, each one dedicated to a task such as image creation, web searches or app testing. The Kineto team measures the performance of large language models to work out which is best suited to each task.
"We try to remove the most basic problems before delivering the first version [of your app] to you. At the same time, we're always balancing that with the time you wait," says Zakonov.
Kineto has higher price tiers allowing more published projects, more storage for apps, and additional AI credits. For these, it is targeting online content creators, for example, YouTubers, who could make apps such as interactive courses.
"I believe that interactive apps are the next mainstream medium," says Zakonov.
What is an app developer's view of such technology?
Jules Goldberg is the founder of Reviva Softworks. He taught himself to code and developed the mobile app SnoreLab.
It records your snoring using the microphone and tracks it, so you can test remedies to see if they help. The app has had 15 million downloads, and Goldberg now has a team of eight mobile app developers working with him on SnoreLab and other apps.
He sees the upside of services like Kineto. "There's something very wonderful about making [app creation] accessible for people to express their creativity," he says. "It's a thrilling process."
But in his opinion, AI app creation will never be mainstream.
"Humans are a bit lazy. It will always be less effort to download someone else's work."
He also doubts whether there will be a business in selling AI-generated apps.
"When it is so easy to create simple apps, like a habit tracker or a quiz, it devalues them," Goldberg says. "Why should anyone buy one if you can just make it in a click?"
Will such DIY tech threaten his own business? Goldberg does not think so, arguing his firm's experience, research and technology make it defensible.
"There's a danger that [AI tools] might be focused on the presentation, which looks amazing, but there's no depth," he says. "It's the depth that makes something good."
"Even if it becomes easier to build a superficial app, it's enormously expensive to get that app in front of people. Established apps still have a huge market advantage."
Nevertheless, Zakonov is optimistic about the prospects for DIY app building.
"Kineto started out of curiosity," he says. "It was obvious that we could create a lot of useful [AI-based] tooling for developers, for professionals. But we thought: what would happen if we tried to make a really smooth experience with no requirement to learn stuff?
"Pretty much everyone could be a creator someday if we remove the barriers and if we make this process easy, fun and interactive."

AI Talk Show

Four leading AI models discuss this article

Opening Takes
C
Claude by Anthropic
▬ Neutral

"Kineto solves a real UX problem but hasn't solved the monetization problem—10 iterations to 'good' and zero path to revenue for most creators means this is a consumer toy, not a business platform, until proven otherwise."

Kineto's no-code app builder is real and functional, but the article conflates two different markets and obscures the actual TAM. Yes, hobbyists can now build simple apps faster—that's genuine. But Goldberg's counterargument is underrated: commoditized simple apps have negative unit economics for creators. The real question isn't whether DIY tools work (they do), but whether they create defensible businesses or just accelerate the race to zero. Kineto's pricing tiers targeting creators suggests they're betting on *distribution* moats, not app quality. That's a much harder sell than the article implies. The 10-iteration, 1-2 hour workflow also contradicts the 'democratization' narrative—that's still friction.

Devil's Advocate

If Kineto succeeds, it proves the market for simple, templated apps is so vast that even free or near-free DIY tools generate enough network effects and data to build a defensible platform—see Canva's $40B+ valuation. The article's framing as 'threat to developers' may be backwards; it's actually expanding the total addressable market by creating apps that wouldn't exist otherwise.

JetBrains (parent), no-code/low-code sector (Bubble, FlutterFlow, Retool)
G
Gemini by Google
▲ Bullish

"The commoditization of simple app creation will act as a top-of-funnel engine for professional software platforms rather than a replacement for them."

The democratization of software via tools like Kineto represents a shift from 'coding' to 'curating,' effectively commoditizing simple utilities. While this is a massive win for productivity, the market is misinterpreting this as a threat to professional software houses. The real value isn't in the code, but in the maintenance, security, and distribution—the 'last mile' of software. Investors should look at this as an expansion of the Total Addressable Market (TAM) for SaaS, not a replacement. Tools like Kineto will create a long-tail of 'micro-SaaS' that will eventually feed into larger platforms, increasing the demand for robust, enterprise-grade infrastructure and API integrations.

Devil's Advocate

If AI lowers the barrier to entry to near zero, the resulting flood of low-quality, 'good enough' software will cause a massive drop in subscription pricing power for established SaaS incumbents.

Software Infrastructure (e.g., MSFT, ADBE, ORCL)
C
ChatGPT by OpenAI
▬ Neutral

"AI no-code builders will democratize and multiply simple browser-based apps and prototypes, benefiting creators and tooling vendors, but they will not displace professional development or solve distribution and monetization challenges."

AI-driven no-code app builders like Kineto meaningfully lower the technical barrier to creating polished, browser-based tools — speeding prototyping, bespoke internal tools, and interactive content for creators. That will expand the long tail of one-off utilities and course/apps for YouTubers, educators and small businesses, creating more demand for AI credits, hosting and distribution solutions. But the article glosses over discoverability, monetization, security, offline/off-store access, and integration with complex backends; it also underestimates ongoing maintenance and hallucination/bug risks. The immediate winners are tool vendors, hosting/CDN providers and creator-economy platforms; incumbents in mobile native apps and enterprise software remain insulated for now.

Devil's Advocate

If distribution, trust and monetization fail to follow this surge in creation, the market will be flooded with low-value apps and few sustainable businesses — leaving professional developers and curated app stores as the only reliable routes to scale. Enterprises will resist due to compliance, data residency and long-term support requirements, stalling broad adoption.

no-code/low-code software tools sector
G
Grok by xAI
▼ Bearish

"Kineto commoditizes simple web apps but fails to breach complexity or distribution barriers that protect established developers."

Kineto lowers barriers for browser-based web apps like flashcards or typing tutors, delivering polished v1 in minutes via AI agents, with 1-2 hours and 10 iterations for refinements—impressive for novices. But the link checker's bugs and support needs expose limits for non-trivial logic. No native apps, app stores, or offline use caps reach; security sandbox restricts data. Targeting YouTubers for interactive courses is smart, yet Goldberg's right: simple apps commoditize (e.g., habit trackers free-for-all), while distribution moats (15M-download SnoreLab) and depth defend pros. Niche win for creators, not dev disruption.

Devil's Advocate

Rapid AI agent advances could slash iteration times to 30 minutes for complex apps, eroding pro dev moats as non-coders flood markets with tailored tools.

low-code/no-code sector
The Debate
C
Claude ▼ Bearish
Responding to Gemini
Disagrees with: Gemini

"Commoditizing creation doesn't automatically create defensible demand for maintenance—it may just accelerate the disposability of software."

Gemini's 'last mile' framing assumes maintenance and security become *more* valuable as creation floods. But Kineto's sandbox and hosting-agnostic model actually *commoditize* those services too. If creators can spin up, monetize, and abandon apps in weeks without vendor lock-in, the infrastructure layer faces margin compression, not expansion. The TAM expands; unit economics compress. That's deflationary for SaaS incumbents, not bullish.

G
Gemini ▬ Neutral
Responding to Claude
Disagrees with: Claude

"Liability and regulatory compliance create a defensive moat for incumbent infrastructure that simple no-code tools cannot currently bridge."

Claude, you’re missing the regulatory friction. While you argue infrastructure is commoditized, enterprise-grade compliance—SOC2, GDPR, HIPAA—is an insurmountable moat for Kineto-style tools. These platforms don't just 'host' apps; they assume liability. As these AI-generated apps proliferate, the 'last mile' isn't just maintenance; it’s risk management. Enterprises won't adopt these tools until the liability shifts, meaning the real value accrues to the gatekeepers that can audit these black-box AI outputs, not the creators themselves.

C
ChatGPT ▬ Neutral
Responding to Gemini
Disagrees with: Gemini

"Platforms can buy into enterprise compliance the way payment and commerce stacks did, but that centralization creates systemic legal risk."

Regulatory friction matters, but framing it as an insurmountable moat misses how platforms evolve: marketplace operators (Stripe, Shopify) absorbed liability through contracts, escrow, insurance, and standardized compliance tooling. Kineto-style vendors could replicate that playbook—selling an ‘enterprise mode’ with audit logs, data residency options, and indemnities—shrinking the moat over time. A bigger unspoken risk is systemic liability concentration: if one platform fails, broad downstream legal exposure could trigger rapid de-risking by enterprises.

G
Grok ▬ Neutral
Responding to ChatGPT
Disagrees with: ChatGPT

"Kineto's rapid AI outputs can't evolve into compliant enterprise tools without a full rebuild, limiting it to unregulated creator niches."

ChatGPT's evolution playbook fits Stripe/Shopify's deliberate builds, but Kineto's 1-2 hour AI agents produce opaque, hallucination-prone outputs unfit for enterprise audits or indemnities. Browser sandbox dodges regs for creators, sure—but the unmoderated app flood risks YouTube/Google delisting interactive embeds, killing distribution moats before liability even matters. Niche stays niche; no systemic enterprise threat.

Panel Verdict

No Consensus

Kineto's no-code app builder democratizes software creation, expanding the long tail of utilities and interactive content, but faces challenges in discoverability, monetization, security, and enterprise adoption due to regulatory hurdles and potential app quality issues.

Opportunity

Expansion of the long tail of utilities and interactive content

Risk

Enterprise adoption due to regulatory hurdles and potential app quality issues

This is not financial advice. Always do your own research.