Claims with a mixed outcome by Ed Zitron
Deadlines are judged as of 2026-10-29. About Ed Zitron and the sources covered.
Claims with a mixed outcome by Ed Zitron (25)
Part of the claim holds and part does not, and the record says which part is which.
Showing 25 of 25 claims, ordered by the date first said, newest first.
Assessment: MixedSaid
Observation
“This was not an “agent” that “went rogue.” It was software doing what software was asked to do, using other bits of software to work out what to do next, all as OpenAI, the company that ran the software, did not appear to have any kind of notification or observability that said “hey man, thousands of LLMs are doing something right now.””
Our reading: The Hugging Face attack was software doing what it was asked, not a rogue agent, and OpenAI appeared to lack any monitoring of thousands of LLMs running.
No horizon Evidence as of
Limitation: The mixed result depends on reading the monitoring criticism as partly about missing evaluation coverage.
Source: AI Is Already In Dangerous Hands
Assessment: MixedSaid
Observation
“For years the AI industry has engaged in endless, vague safety theater about the “risks” of AI, all while peddling software that is actively harmful and unreliable.”
Our reading: For years the AI industry has engaged in vague safety theater while selling harmful, unreliable software.
No horizon Evidence as of
Limitation: The result depends on reading safety theater as the absence of tangible mitigation, following his surrounding explanation.
Source: AI Is Already In Dangerous Hands
Assessment: MixedSaid
Causal claim
“Its software lies because it doesn’t actually know anything and generates answers probabilistically”
Our reading: AI software lies because it doesn't actually know anything and generates answers probabilistically.
No horizon Evidence as of
Limitation: The result depends on reading knowledge as accessible factual information.
Source: The AI Hater's Manifesto
Assessment: MixedSaid
Observation
“Anthropic has not created recursive self-improvement, nor has it done much more than create a very large language model that gets higher benchmarks in tests built for large language models, wrapped in a veneer of mysticism and panic-hype built to scare organizations in paying them to use it.”
Our reading: Anthropic has not created recursive self-improvement, nor has it done much more than create a very large language model that scores higher on benchmarks built for large language models, wrapped in mysticism and panic-hype meant to scare organizations into paying to use it.
No horizon Evidence as of
Limitation: The recursive self-improvement evidence is a company disclosure from June 4, shortly before the post.
Source: AI's Brokenomics
Assessment: MixedSaid
Causal claim
“This issue bubbled to the surface in the last few months because Anthropic and OpenAI both quietly moved all of their enterprise customers to token-based billing in Q1 2026, and because these enterprise customers are run by Business Idiots with no connection to actual work, CEOs encouraged (or actively incentivized) their workers to use AI as much as possible, in some cases even making one’s AI use a KPI that could cost them their job.”
Our reading: The AI cost problem surfaced in the last few months because Anthropic and OpenAI quietly moved all their enterprise customers to token-based billing in Q1 2026, and because CEOs had encouraged or incentivized workers to use AI as much as possible, in some cases making AI use a job-threatening KPI.
No horizon Evidence as of
Limitation: Anthropic's migration record does not establish OpenAI's Q1 customer coverage [e1].
Source: AI Doesn't Have ROI
Assessment: MixedSaid
Causal claim
“Because AI data centers are so capital-intensive, they require endless amounts of risky debt, but that risk allows private credit to take investments from insurance and pension funds desperate for yield, and because everybody involved is a Business Idiot, nobody has actually thought about what happens if these things don’t work out.”
Our reading: Because AI data centers are so capital-intensive they need endless risky debt, which lets private credit take money from yield-hungry insurance and pension funds, and because everyone involved is a Business Idiot, nobody has actually thought about what happens if these projects don't work out.
No horizon Evidence as of
Limitation: The contradictory evidence addresses absence of consideration, not whether preparations were sufficient.
Source: Revenge of The Business Idiot
Assessment: MixedSaid
Causal claim
“while prices for some models might have come down, newer “reasoning” models burn way more tokens, which means the cost of inference has, somehow, gotten higher over time.”
Our reading: Although prices for some models might have come down, newer reasoning models burn far more tokens, so the cost of inference has gotten higher over time.
No horizon Evidence as of
Limitation: The studies do not establish an industry-wide trend across production workloads or providers' internal inference costs [e8, e9].
Source: AI's Economics Don't Make Sense
Assessment: MixedSaid
Observation
“This does not really make sense given the fact that all support documents and the Claude website reflect that Pro users do not have access to Claude Code.”
Our reading: Anthropic's claim that this was a small test on about 2% of new signups does not fit the fact that all support documents and the Claude website showed Pro users without Claude Code access.
No horizon Evidence as of
Limitation: The counterexample relies on contemporaneous firsthand reporting because a sufficiently complete archived product page was not recovered.
Assessment: MixedSaid
Causal claim
“A coding model relies heavily on chain-of-thought reasoning, which means that despite the cost of tokens coming down (which does not mean the price of providing them has decreased, it’s a marketing move), models are using far, far more tokens, increasing costs across the board.”
Our reading: Coding models rely heavily on chain-of-thought reasoning and so use far more tokens, which raises costs across the board even though token prices have fallen; lower token prices do not mean the cost of providing tokens has fallen, and are a marketing move.
No horizon Evidence as of
Limitation: The decisive coding evidence is company-reported.
Source: The Subprime AI Crisis Is Here
Assessment: MixedSaid
Causal claim
“Deep Impact, my analysis of the DeepSeek situation and why it scared the American AI industry (clue: it's cost-related and nothing to do with "national security").”
Our reading: DeepSeek scared the American AI industry because of cost, not because of national security.
No horizon Evidence as of
Limitation: The market report does not establish individual investors' or executives' private motives [e1].
Source: 2025, A Retrospective
Assessment: MixedSaid
Observation
“Large Language Models are limited, AI agents are a pipedream and simply do not work, AI-powered products are unreliable and coding LLMs make developers slower, and the cost of inference — the way in which a model produces its output — keeps going up.”
Our reading: LLMs are limited, AI agents do not work, AI products are unreliable, coding LLMs make developers slower, and inference costs keep rising.
No horizon Evidence as of
Limitation: The coding result is specific to experienced developers working in familiar repositories.
Assessment: MixedSaid
Causal claim
“They don’t even speed up engineers — they actually make them slower!”
Our reading: Coding LLMs do not speed up engineers and actually make them slower.
No horizon Evidence as of
Limitation: The trials should not be pooled into an industry-wide estimate without accounting for their different settings.
Source: The Case Against Generative AI
Assessment: MixedSaid
Forecast
“So what, OpenAI makes $12.7 billion this year, but loses $14 billion, what does that mean to you, exactly?”
Our reading: OpenAI will make about $12.7 billion in revenue in 2025 but lose about $14 billion.
Deadline passed Evidence as of
Limitation: The numerical tolerance and loss interpretation were supplied by the evaluator.
Source: Make Fun Of Them
Assessment: MixedSaid
Observation
“Converting cryptocurrency mining data centers to HPC data centers is effectively starting from scratch. There is no easy or logical way to repurpose a bitcoin miner for AI.”
Our reading: Bitcoin mining infrastructure cannot easily be repurposed for AI; conversion is effectively starting from scratch.
No horizon Evidence as of
Limitation: This is a split assessment of hardware replacement and facility reuse.
Source: CoreWeave Is A Time Bomb
Assessment: MixedSaid
Observation
“The one "big thing" that they've been able to do is to use "reasoning" to make the Large Language Models "think" (they do not have consciousness, they are not "thinking," this just means using more tokens to answer a particular question and having multiple models check the work), which mostly results in them being a bit more accurate when generating an answer, but at the expense of speed and cost.”
Our reading: The one big thing AI model companies have been able to do is 'reasoning', which means using more tokens and multiple models checking the work rather than actual thinking, and mostly results in models being a bit more accurate at the expense of speed and cost.
No horizon Evidence as of
Limitation: The performance results are developer-reported and concentrated on reasoning tasks.
Source: The Generative AI Con
Assessment: MixedSaid
Observation
“The launch of its "Operator" "agent" was a joke — a barely-functional product that is allegedly meant to control your computer and take distinct actions, but doesn't seem to work.”
Our reading: OpenAI's Operator agent is barely functional and does not seem to work.
No horizon Evidence as of
Limitation: Successful simple tasks do not establish dependable completion of harder tasks.
Source: Deep Impact
Assessment: MixedSaid
Observation
“They're running out of training data, and the actual capabilities of these models were peaking as far back as March 2024.”
Our reading: LLM developers are running out of training data, and model capabilities had been peaking since March 2024.
No horizon Evidence as of
Limitation: Support for data scarcity depends on the ordinary human-text reading and continuation of scaling trends.
Source: Deep Impact
Assessment: MixedSaid
Observation
“The hallucination problem is one that is nowhere closer to being solved — and, at least with the current technology — may never go away, and it makes it a non-starter for a great many business tasks, where you need a high level of reliability.”
Our reading: The hallucination problem is nowhere closer to being solved, may never go away at least with current technology, and makes generative AI a non-starter for a great many business tasks that need high reliability.
No horizon Evidence as of
Limitation: The supporting finding concerns legal research and cannot establish the full breadth of great many.
Source: Godot Isn't Making it
Assessment: MixedSaid
Observation
“far too many simply ran exactly what Musk said and acted as if he’d proven that these were actual robots walking around and doing things on their own accord, as opposed to what was blatantly obvious: that they were controlled remotely by human operators.”
Our reading: The Tesla Optimus 'AI-powered' robots shown at the We, Robot event were controlled remotely by human operators rather than acting autonomously.
No horizon Evidence as of
Limitation: Evidence comes from news reporting of footage and participant accounts, rather than controller logs or a directly readable company disclosure.
Assessment: MixedSaid
Observation
“ChatGPT Plus has no real moat, little product differentiation (outside of its advanced voice mode, which Meta is already working on a competitor to), and increasing commoditization from other models, open source platforms, and even on-device models (like on Copilot+-powered PCs).”
Our reading: ChatGPT Plus has no real moat, little product differentiation apart from advanced voice mode (which Meta is already working to match), and faces increasing commoditization from other models, open-source platforms and on-device models.
No horizon Evidence as of
Limitation: The mixed judgment turns on treating the measured reasoning advantage as meaningful product differentiation.
Source: OpenAI Is A Bad Business
Assessment: MixedSaid
Observation
“OpenAI failed to make a more efficient model for Microsoft late last year, and while GPT-5 is meant to be “materially better,” it isn’t obvious what “better” means when GPT-4 performs worse at some tasks than its predecessor.”
Our reading: OpenAI failed to build a more efficient model for Microsoft in late 2023, and GPT-4 performs worse than its predecessor at some tasks.
No horizon Evidence as of
Limitation: Arrakis evidence is secondary and traces to one underlying report.
Source: Expectations Versus Reality
Assessment: MixedSaid
Observation
“Every single story about artificial intelligence seems to begin with “well, AI can’t do much yet, but…” and then lean into sloppy science fiction that masks that these “autonomous agents” are effectively chatbots that can answer questions based on a document, tech that’s existed for a decade or more.”
Our reading: So-called autonomous agents are effectively chatbots answering questions from documents, a decade-old technology.
No horizon Evidence as of
Limitation: A research agent does not establish the quality of every commercial product.
Source: Expectations Versus Reality
Assessment: MixedSaid
Causal claim
“As I've discussed before, today's models are rife with hallucinations that make integrating them actively dangerous for many businesses, and the only way to have fewer hallucinations is to give the models more compute power to process and analyze data, a dwindling resource that AI is actively helping to destroy.”
Our reading: Today's models are rife with hallucinations that make integrating them actively dangerous for many businesses, and the only way to reduce hallucinations is to give models more compute power to process and analyze data, which is a dwindling resource that AI is actively helping to destroy.
No horizon Evidence as of
Limitation: The counterexample concerns deployed processing capacity, not total development computation.
Source: Bubble Trouble
Assessment: MixedSaid
Causal claim
“Amazon already plans to heavily integrate generative AI into Alexa, and Apple is planning the same with Siri, all to show the world that they're "innovative" and "pushing boundaries" without being able to explain how.”
Our reading: Amazon plans to heavily integrate generative AI into Alexa and Apple plans the same for Siri, both to show the world they are "innovative" and "pushing boundaries" without being able to explain how.
No horizon Evidence as of
Limitation: The adverse finding depends on reading 'explain how' as articulating concrete benefits, rather than satisfying the speaker's judgment of innovation.
Source: Software Has Eaten The Media
Assessment: MixedSaid
Observation
“While it may seem strange that he’d drop his metaverse idea at the first sign of something more perceptually interesting to Wall Street, he has a proven track record of chasing other people’s ideas and failing to compete.”
Our reading: Zuckerberg dropping his metaverse idea at the first sign of something more perceptually interesting to Wall Street, which may seem strange, is consistent with his proven track record of chasing other people's ideas and failing to compete.
No horizon Evidence as of
Limitation: The failed-product examples rely on contemporary technology reporting.
Source: Fire Mark Zuckerberg
No claims match. Remove a filter or change the search, or clear everything.