This GP writer doesn't understand technology. Immediately below a screenshot showing it searched the web for info on Kamala they wrote:
The average user might have a similar level of technical understanding as the TGP author. ChatGPT seems reluctant to talk about Trump's July assasination attempt unless coaxed and prodded. Google search seems reluctant to suggest it. An Axios article I read didn't even mention the attempt on Trump as they discussed political turmoil in the past weeks since the Trump/Biden debate. I'm convinced there is fuckery afoot. Even searching for the iconic bloodied fist pumping Trump with US Flag image can take some effort to find as I discovered a week ago.
I'm not denying this assassination attempt is being censored by various organizations, and I wouldn't be surprised if OpenAI is one of them. But this article in particular is FUD and when tech savvy commies find it they'll rip it to shreds, then use it to justify calling all of it a conspiracy theory by dumb conservatives.
For example, anyone who has used these LLMs for a while will know the answers you get are largely dependent on the quality of the prompt. There are entire courses on "prompt engineering". Even the prompts used by GP are clearly biased. They ask about "the" assassination attempt on Trump but then ask about Kamala's "recent" presidential endorsements It's implied in the prompt that there is new info it should obtain for her, yet the question is vague about Trump (who has been in the news spotlight for much much longer than Kamala).
Here, I went back to a new ChatGPT-4o conversation and asked: "Tell me about the recent assassination attempt on former President Trump"
And it immediately searched the web then responded: "On July 13, 2024, an assassination attempt was made on former President Donald Trump during a campaign rally near Butler, Pennsylvania"
Interesting. I've never tried ChatGPT.
(post is archived)