We had a big company wide town hall yesterday where the CEO went over the plan for the next few fiscal years, including specific goals, figures, etc. As I've mentioned, the company has been plugging AI into everything lately. Out of curiosity, since it transcribes in real time, I asked it to provide a summary of the hour long meeting. It took a good minute or so, but it produced a frighteningly accurate and concise summary of all the main points, including citing specific numbers. There was even a little blurb at the end about the safety compliance officer that opened the meeting and gave a brief set of recommendations. It was a bit unsettling.
For fun, I ran that text through it to clean it up. Here's the result:
“Yesterday, we held a company-wide town hall where the CEO outlined the plan for the next few fiscal years, including specific goals and figures. As I mentioned earlier, the company has been integrating AI into various aspects of our operations. Out of curiosity, during the hour-long meeting, I asked the AI to provide a real-time summary. Surprisingly, it delivered an incredibly accurate and concise overview, complete with specific numbers. At the end, the safety compliance officer shared a brief set of recommendations. The whole experience left me feeling a bit unsettled.”
It missed the context of the safety officer, but yeah.
Anyone essentially saying “I could do that better” unironically should never again be considered for a commentary panel. Whether or not the claim is true is beside the point — it brings no value to the viewer either way.
I've been using our AI tool at work for a few months now. It can be a real time saver for things like summarizing interview transcripts and organizing notes into an outline. But I have to rewrite pretty much everything it spits out because it has no style.
Yeah, comparing the two versions of the CoC, the cleaned-up text is mechanically correct but bland. Also, I question the decision to reframe terrifying as surprising.
Our tool has a sliding scale setting for precision versus creativity. Since it is trained on the full internet I almost always favor more precision since I want maximum accuracy and minimal hallucinations. Full creative mode can give you some hilarious results.
If you can use the tool to take detailed notes for a meeting, it is a definite win. Nobody likes being the notetaker.
is anyone doing anything for this year's HOF ceremony?
A couple friends are heading there and I think I would as well, but the ceremony is the day we’re flying back from Iceland. Bad timing!
and so it goes...
Underrated comment. I wish there was a dollar system.
The first three episodes of a podcast detailing what is, and likely will always be, the biggest scandal in curling are out. Its a pretty fun story and had ramifications on the Madison club that lead to Matt Hamilton sending the greatest club wide email in the history of the club.
I binged all three last night. Fascinating story.
And the accents - they all sound like Letterkenny/Shorsey characters
And its only just getting good, still three more episodes to go.
I was almost sort of mad at you that you shared it before all the episodes were released hahaha
Did you find that email?
I haven't yet. I may have deleted it but I'm sure one of my friends has it somewhere.
I use chatbotz to first draft lots of things. I also decided that my next chapter needed to not be related to something that the AI overlords could do from their virtual throne. We’ll see if anyone wants their paper preserved or if we’re all just too amused by our personal ai friends.
TNT should show this highlight reel.
Anyone essentially saying “I could do that better” unironically should never again be considered for a commentary panel. Whether or not the claim is true is beside the point — it brings no value to the viewer either way.
I've been using our AI tool at work for a few months now. It can be a real time saver for things like summarizing interview transcripts and organizing notes into an outline. But I have to rewrite pretty much everything it spits out because it has no style.
Yeah, comparing the two versions of the CoC, the cleaned-up text is mechanically correct but bland. Also, I question the decision to reframe terrifying as surprising.
Our tool has a sliding scale setting for precision versus creativity. Since it is trained on the full internet I almost always favor more precision since I want maximum accuracy and minimal hallucinations. Full creative mode can give you some hilarious results.
If you can use the tool to take detailed notes for a meeting, it is a definite win. Nobody likes being the notetaker.
is anyone doing anything for this year's HOF ceremony?
A couple friends are heading there and I think I would as well, but the ceremony is the day we’re flying back from Iceland. Bad timing!
and so it goes...
Underrated comment. I wish there was a dollar system.
The first three episodes of a podcast detailing what is, and likely will always be, the biggest scandal in curling are out. Its a pretty fun story and had ramifications on the Madison club that lead to Matt Hamilton sending the greatest club wide email in the history of the club.
I binged all three last night. Fascinating story.
And the accents - they all sound like Letterkenny/Shorsey characters
And its only just getting good, still three more episodes to go.
I was almost sort of mad at you that you shared it before all the episodes were released hahaha
Did you find that email?
I haven't yet. I may have deleted it but I'm sure one of my friends has it somewhere.
I use chatbotz to first draft lots of things. I also decided that my next chapter needed to not be related to something that the AI overlords could do from their virtual throne. We’ll see if anyone wants their paper preserved or if we’re all just too amused by our personal ai friends.