AI projects keep dying while hype stays alive

The expanding AI graveyard

TechCrunch catalogued a growing list of AI ventures that have folded or missed expectations, from Apple’s repeatedly delayed Siri upgrades to OpenAI’s chaotic “super app” launch. The list reads like a roll call of overpromised projects that never delivered functional products.¹²³

The article names Apple, OpenAI, and several unnamed startups as examples of initiatives that stalled despite public hype. It highlights Siri’s delayed AI improvements and OpenAI’s attempt to bundle multiple services into a single app that quickly ran into performance and user‑experience problems.^1456278 The pattern is clear: big names launch ambitious AI features, then retreat or retool when the technology fails to meet market expectations.

Why the hype outpaces delivery

The tech press often treats AI as a magic bullet, but the reality is far messier. Prominent thinkers like Eliezer Yudkowsky, praised by Sam Altman, sell a vision of “perfect health” and “immortality” through artificial general intelligence. Those promises inflate expectations faster than research can substantiate them.

Yudkowski’s transhumanist ambitions trace back to centuries‑old alchemical quests for the Philosopher’s Stone. The modern echo appears in venture capital decks that promise limitless abundance while ignoring the incremental nature of model training, data curation, and compute limits. When projects like OpenAI’s super app stumble, the fallout is not just a missed product launch; it erodes confidence in the broader narrative that AI will solve humanity’s biggest problems.⁸

Industry insiders note that the gap between hype and deliverable features creates a “graveyard” effect. Companies pour resources into flagship AI announcements, then quietly shelve the work when performance metrics fall short. The result is a cycle of publicity, disappointment, and reallocation of engineering talent to more incremental improvements.

Infrastructure and the platform promise

Amazon Web Services continues to market its machine‑learning ecosystem as a turnkey solution for developers. A 2018 keynote by Dan Romuald Mbanga at JupyterCon highlighted AWS’s API services and learning resources on the O’Reilly platform. The presentation emphasized skill‑building and integration, but it stopped short of guaranteeing that every AI startup can scale on the infrastructure.

The AWS pitch assumes that access to powerful compute and curated datasets will translate into successful products. In practice, many AI ventures still falter because the underlying models lack robustness or because the business case for their deployment is weak. The contrast between AWS’s confident platform narrative and the reality of failed AI projects underscores a deeper mismatch between infrastructure promises and product viability.

Regulatory and mythic underpinnings

Policymakers have begun to treat AI as a national‑security issue. The Biden administration, for example, has bet on the premise that artificial general intelligence is imminent, while previous administrations and foreign allies have framed AI as a tool for geopolitical leverage. These stances amplify the mythic status of AI, echoing the historic allure of alchemy described by Rudolf II and later dramatized in literature.

The mythic framing fuels investment cycles that prioritize headline‑grabbing claims over rigorous validation. When the promised breakthroughs fail to materialize, the fallout extends beyond corporate balance sheets to shape public perception and regulatory agendas. The AI graveyard therefore becomes a barometer for how quickly myth outpaces measurable progress.

What to watch

Future tracking should focus on whether major AI players shift from grand announcements to transparent roadmaps that include measurable milestones. Pay attention to any formal disclosures from companies like Apple and OpenAI about the status of delayed features or shelved products.

Also monitor policy developments from the White House and foreign governments that tie funding or regulation to the achievement of AGI milestones. If those policies begin to require independent audits or performance benchmarks, the industry may finally align hype with accountable delivery.