HeyJaredAI Unpacked
Every video weighs one claim behind a named AI product — sourced numbers, self-reported vs verified, no winner declared.
- Indexed videos, last 90 days
- 202
- Latest publication
- Sep 30, 2026
- Audience
- ~42 subscribers
- Earliest in this view
- Jul 3, 2026
Latest videos
Was The Cashierless Store Actually Run By 1,000 People? (opens the original)
Read excerpt
"Amazon's Just Walk Out stores were never AI - they were 1,000 people in India watching you shop." It is one of the most repeated AI debunkings of the decade, and in the form everyone shared it, it is wrong. This video puts it on trial. There was no room of workers watching a live feed and ringing up baskets. The system was genuine computer vision - ceiling cameras, shelf weight sensors, multi-person tracking through occlusion in a crowded store, one of the least forgiving deployment environment
A Web Page Can Give Your AI Orders - And It Obeys (opens the original)
Read excerpt
Hide a sentence on a web page that no human will ever see, and an AI assistant reading that page on your behalf will not treat it as text. It will treat it as an instruction. This is indirect prompt injection, and it is not a bug - it is a consequence of a design property: a language model has one input channel, so your instructions and the content it reads arrive as the same kind of token, with nothing structurally marking which is which. This episode audits the numbers, because the public conv
How long flawed research affects classrooms #Shorts (opens the original)
Read excerpt
A meta-analysis published in May 2025 reported that ChatGPT produced a large positive effect on student learning performance (Hedges' g = 0.867). Narration and some visuals are AI-assisted. ▶ Subscribe — we check the claims behind every AI launch. #chatgptineducation #retractedstudy #aiinclassrooms ai education research ai explained ai fact check ai in classrooms ai learning outcomes chatgpt in education chatgpt students does chatgpt help students education technology effect size meta analysis r
ChatGPT: What is a meta-analysis? #Shorts (opens the original)
Read excerpt
A meta-analysis published in May 2025 reported that ChatGPT produced a large positive effect on student learning performance (Hedges' g = 0.867). Narration and some visuals are AI-assisted. ▶ Subscribe — we check the claims behind every AI launch. #chatgptineducation #retractedstudy #aiinclassrooms ai education research ai explained ai fact check ai in classrooms ai learning outcomes chatgpt in education chatgpt students does chatgpt help students education technology effect size meta analysis r
ChatGPT Invented Police Witnesses In A Murder Appeal (opens the original)
Read excerpt
On September 11, 2026, the New Mexico Supreme Court held a defense attorney in contempt, fined him $5,000, struck his brief and ordered a murder appeal to restart from the beginning. The brief quoted testimony from police witnesses who never testified, including a fictional detail that the shooter wore dark pants and a white shirt. The lawyer said he had used ChatGPT to summarize the trial proceedings and had not understood how far these systems can fabricate. This episode does three things. Fir
Publishing over time
Last 90 days. Choose a month to open its work.
Recurring subjects
Named in the text we hold. One piece can cover several.
Audience
~42 subscribers
Measured Sep 19, 2026
Source's subscribers, not the number who saw an individual piece.
About this data
Counts cover the work we have indexed. Tone needs enough text and a confident classification. Excerpts and episode notes are not full articles or transcripts.
Identity or attribution wrong? Suggest a correction.