Subprime Intelligence
We recorded 9 claims from this post. 9 can’t be checked. 15 more are held. Read the post at its source.
Claims from this post (9)
In the order they appear in the post. Each quotation is the speaker’s words in this post. A label such as “Supported” or “Can’t be checked” is the claim’s current assessment; its page has the evidence and the reasons.
Assessment: Can’t be checkedSaid
Observation
“Sora's outputs can mimic real-life objects in a genuinely chilling way, but its outputs — like DALL-E, like ChatGPT — are marred by the fact that these models do not actually know anything.”
Our reading: Generative models like Sora, DALL-E and ChatGPT do not actually know anything.
No horizon
Assessment: Can’t be checkedSaid
Forecast
“These things are not "learning," or "understanding," or even "intelligent" — they're giant math machines that, while impressive at first, can never assail the limits of a technology that doesn't actually know anything.”
Our reading: Generative AI models are not learning, understanding or intelligent and can never overcome the limits of a technology without knowledge.
No horizon
Assessment: Can’t be checkedSaid
Observation
“I believe artificial intelligence companies deeply underestimate how perfect the things around us are, and how deeply we base our understanding and acceptance of the world on knowledge and context.”
Our reading: Zitron believes AI companies deeply underestimate how perfect the things around us are and how deeply we base our understanding and acceptance of the world on knowledge and context.
No horizon
Qualifier: I believe
Assessment: Can’t be checkedSaid
Observation
“As I discussed in my last newsletter, media entities are increasingly normalizing their content to please search engine algorithms, and the jobs that involve pooling affiliate links and answering where you can watch the Super Bowl are very much at risk.”
Our reading: Media jobs producing SEO-driven generic content, like affiliate-link pooling, are very much at risk from generative AI.
No horizon
Assessment: Can’t be checkedSaid
Forecast
“I don't argue it will, at least not in such a way that anybody's lives will be improved.”
Our reading: Zitron doubts that generative AI, a product that can only sort-of-kind-of do something, will create trillions of dollars of economic value, at least not in a way that improves anybody's lives.
No horizon
Qualifier: at least not in such a way that anybody's lives will be improved
Assessment: Can’t be checkedSaid
Recommendation or value judgment
“I'd argue that a generative AI's inability to reliably generate stuff is an existential threat that should have smothered these companies early in their lives.”
Our reading: Zitron argues that generative AI's inability to reliably generate output is an existential threat that should have smothered AI companies like OpenAI and Anthropic early in their lives.
No horizon
Qualifier: I'd argue
Assessment: Can’t be checkedSaid
Observation
“Google's investment in Anthropic was made in the hopes that it’d see a similar revenue multiplier, and I'd argue Amazon's was made in the same vein — though it was too late to force Anthropic to use AWS as their preferred vendor.”
Our reading: Google's investment in Anthropic was made hoping for a revenue multiplier like Microsoft's, and Zitron argues Amazon's was made in the same vein, though too late to force Anthropic to use AWS as its preferred vendor.
No horizon
Qualifier: I'd argue (as to Amazon)
Assessment: Can’t be checkedSaid
Observation
“I consider myself relatively tuned into the tech ecosystem, and I read every single tech publication regularly, yet I'm struggling to point to anything that generative AI has done other than reignite the flames of venture capital.”
Our reading: Zitron, who says he reads every tech publication regularly, is struggling to point to anything generative AI has done other than reignite venture capital.
No horizon
Qualifier: in his own assessment
Assessment: Can’t be checkedSaid
Observation
“Yet even with thousands of headlines, billions of dollars of investment, and trillions of tokens run through various large language models, there are no essential artificial intelligence use cases, and no killer apps outside of non-generative assistants like Alexa that are now having generative AI forced into them for no apparent reason.”
Our reading: Despite thousands of headlines, billions of dollars of investment and trillions of tokens, there are no essential AI use cases and no killer apps apart from non-generative assistants like Alexa.
No horizon
Scope: outside of non-generative assistants like Alexa
Held from this post (15)
These records did not pass every check, so nothing about them is published. Held is not a finding.
- The two reviews read the claim differently
- The two reviews reached different assessments. Our reading, not assessed: Generative AI models are unprofitable, and Altman's desperation for a new chip company or energy breakthrough suggests they will only become more unprofitable as they generate more revenue.
- The two reviews read the claim differently
- The two reviews reached different assessments. Our reading, not assessed: Even if the costs were not prohibitive, one cannot make a watchable movie, TV show or even commercial out of AI video outputs that are not consistent from clip to clip, because even the smallest errors are outright repulsive to viewers.
- The two reviews read the claim differently
- The two reviews reached different assessments. Our reading, not assessed: The inevitable outcome is that generative AI models are not reliable enough to actually replace anyone, and those who have experimented with replacing people have found themselves deeply embarrassed.
- The two reviews reached different assessments. Our reading, not assessed: Given reported gross margins of 50-55% versus 75-90% for modern software companies, the raw infrastructure firms that let startups integrate AI are not particularly healthy businesses and keep far less of their money.
- The two reviews reached different assessments. Our reading, not assessed: Anthropic and OpenAI are not really at risk, because they have become an important part of the tech ecosystem as the tail of a very hungry snake that feeds big tech.
- The two reviews read the claim differently
- The two reviews read the claim differently
- The two reviews read the claim differently
- The two reviews reached different assessments. Our reading, not assessed: Microsoft's, Amazon's and Google's 'investments' in AI companies really created guaranteed revenue streams, because the recipients are effectively obliged to spend the money on the investors' own services.
- The two reviews read the claim differently
- The two reviews read the claim differently
- The two reviews reached different assessments. Our reading, not assessed: Hallucinations in generative AI cannot be fixed, because the models predict from tags in a dataset, which they might do well but can never do flawlessly or reliably.