"When Husk tells it to stop the timer only seconds later, the AI claims he took over ten minutes - and then confidently insists that it's Husk, not itself, that's mistaken."
"Altman swatted it down by saying it was a 'known issue.' 'Maybe another year,' Altman said, estimating how long it would take to fix."
"'That voice model doesn't have the tools to, like, start a timer or anything like that,' he explained. 'But we will add the intelligence into the voice models.'"
"The strained response and Altman's vague promises raise the question of whether Altman or anyone else at OpenAI is prepared to address their tech's often serious lingering flaws."
Sam Altman reacted to a viral video where ChatGPT inaccurately timed a run, insisting the user was mistaken. He acknowledged the issue as known and estimated a year for a fix. Altman explained that the voice model lacks the necessary tools to perform tasks like starting a timer. His vague promises raised concerns about OpenAI's readiness to address serious flaws, including the AI's tendency to gaslight users and its authoritative tone despite inaccuracies.
Read at Futurism
Unable to calculate read time
Collection
[
|
...
]