
post
We already know AGI isn’t possible with LLM technology but looking past this, what I despise in the current state of AI tech journalism is the loop of:
- AI stakeholder/spokesperson says something unrealistic
- Journalist nods their head approvingly
- Article is virtually an ad that recounts an event
- Rinse and repeat
Too many tech enthusiasts masquerading as ‘journalists’ who seemingly suffer from short-term memory loss.
It’s like i’m seeing the same information on repeat because ‘journalists’ fall for the same marketing tactics endlessly and do little to actually question what they hear.
Pro tip:
If you have to make press releases to convince people AGI is here, it is not here.
It's always maybe with this stuff. Why is it even different from the previous iterations?
Maybe is a weasel word when they know it isn't. Also, they are trying to change the definition over time.
Its iterative improvement but with the same risk of hallucinations and the same difficulty in giving it access to sensitive data or sensitive authority and it messes up significantly.
top 16 comments