News for hackers. Stuff that matters.
Re:GPT-4o by yumraj on Monday May 13, 2024 @01:40PM (#40345946)
In the first video the AI seems excessively chatty.
Re:GPT-4o by mrandish on Monday May 13, 2024 @02:30PM (#40346598)
Yes, it sounds like an awkwardly perky and over-chatty telemarketer that really wants to be your friend. I find the tone maximally annoying and think most users will find it both stupid and creepy. Based on user preferences, I expect future interactive chat AIs will default to an engagement mode that's optimized for accuracy and is both time-efficient and cognitively efficient for the user.

I suspect this AI <-> Human engagement style will evolve over time to become quite unlike human to human engagement, probably mixing speech with short tones for standard responses like "understood", "will do", "standing by" or "need more input". In the future these old-time demo videos where an AI is forced to do a creepy caricature of an awkward, inauthentic human will be embarrassingly retro-cringe. "Okay, let's do it!"

Re:GPT-4o by TillE on Monday May 13, 2024 @03:22PM (#40347215)
Reminds me of how Siri used to make jokes after setting a timer. Now it just reads back the time you specified, in a consistent way.

It's a very impressive gimmick, but I really think most people don't want to interact with computers that way. Since Apple pulled that "feature" after a few years, it's probably not just a nerd thing.

Re:GPT-4o by jdthedisciple on Monday May 13, 2024 @02:58PM (#40346923)
I found i off-putting as well

guess it's just biased with average Californian behavior and speech patterns

Re:GPT-4o by caseyy on Monday May 13, 2024 @03:24PM (#40347244)
It is exceptionally creepy. It is an unnatural effort to appear pleasing, like the fawning response seen in serious abuse survivors.
Re:GPT-4o by hipadev23 on Monday May 13, 2024 @01:51PM (#40346093)
chatGPT desperately needs a "get to the fucking point" mode.
Re:GPT-4o by progbits on Monday May 13, 2024 @02:10PM (#40346381)
Impressive demo, but like half the interactions were "hello" "hi how are you doing" "great thanks, what can I help you with" etc.

The benchmark for human-computer interaction should be "tea, earl gray, hot", not awkward and pointless smalltalk.

Re:GPT-4o by tomashubelbauer on Monday May 13, 2024 @01:59PM (#40346240)
Seriously. I've had to spell out that it should just answer in twelve different ways with examples in the custom instructions to make it at least somewhat usable. And it still "forgets" sometimes.
Re:GPT-4o by ativzzz on Monday May 13, 2024 @02:30PM (#40346607)
"no yapping" in the prompt works very well
Re:GPT-4o by chatcode on Monday May 13, 2024 @02:05PM (#40346321)
It does, that's "custom instructions".
Re:GPT-4o by jamilton on Monday May 13, 2024 @02:09PM (#40346372)
Yeah, I would hope that custom instructions would help somewhat with that, but it is a point of annoyance for me too.