 |
Post Ay2lt2moXunpztducK by [email protected] |
 |
More posts by [email protected] |
|
Post #Axyvn8ec2Tq3Mqn0wy by [email protected] |
|
0 likes, 0 repeats |
|
|
|
"How can you get more accurate answers from an LLM?” is like asking &quo… |
|
|
|
Post #Axyvn8kzekwVgdm6tc by [email protected] |
|
0 likes, 1 repeats |
|
|
|
@thomasfuchs not quite. I've found using language like "using only wel… |
|
|
|
Post #Axywp19N0dP8czdUg4 by [email protected] |
|
0 likes, 0 repeats |
|
|
|
@codinghorror Just like with other generative outputs, in summaries LLMs often … |
|
|
|
Post #Axywp1GSaH4kyyx9jE by [email protected] |
|
0 likes, 0 repeats |
|
|
|
@codinghorror You can’t hack yourself to magically change how they work—the… |
|
|
|
Post #Axywp1M8FBc3GZbgZM by [email protected] |
|
0 likes, 1 repeats |
|
|
|
@thomasfuchs right, it always requires centaur -- human review. But I ardently … |
|
|
|
Post #Axywxqij9jhkt3NxEO by [email protected] |
|
0 likes, 1 repeats |
|
|
|
@thomasfuchs I am just one data point, but I look VERY CLOSELY at these outputs… |
|
|
|
Post #AxyxCAwwjyiFgIrqgC by [email protected] |
|
0 likes, 0 repeats |
|
|
|
@codinghorror You perhaps do. But 99.99% of people using these tools don’t. |
|
|
|
Post #AxyxCB2yNZX7yzgf4a by [email protected] |
|
0 likes, 1 repeats |
|
|
|
@thomasfuchs that's the thing to warn people about. Sourcing (label the ing… |
|
|
|
Post #AxyxFlvDjeppDWm7Rw by [email protected] |
|
0 likes, 1 repeats |
|
|
|
@thomasfuchs basically it is a research assistant who is enthusiastic and never… |
|
|
|
Post #AxyxgDNaGuO7u7ePi4 by [email protected] |
|
0 likes, 0 repeats |
|
|
|
@codinghorror @thomasfuchs I wouldn’t even trust an LLM to correctly parse �… |
|
|
|
Post #AxyxgDUJrrmAF0nnCy by [email protected] |
|
0 likes, 1 repeats |
|
|
|
@wbftw @thomasfuchs in my experience this is parsed correctly. It's backlin… |
|
|
|
Post #AxyyEbs7iPc7iEKuJ6 by [email protected] |
|
0 likes, 1 repeats |
|
|
|
@thomasfuchs reminder that we predict the weather (quite accurately, nowadays, … |
|
|
|
Post #AxyzOlILdTMxms0Vgu by [email protected] |
|
0 likes, 0 repeats |
|
|
|
@codinghorror two things can be bad at the same time ¯\_(ツ)_/¯ |
|
|
|
Post #AxyzOlONH4Bq5YpK5I by [email protected] |
|
0 likes, 1 repeats |
|
|
|
@thomasfuchs the median has value, though, because all the data points to it be… |
|
|
|
Post #AxyzWXWWil6hOQLdi4 by [email protected] |
|
0 likes, 0 repeats |
|
|
|
@codinghorror Communication needs to be precise because when you conflate hypot… |
|
|
|
Post #AxyzWXcuL2D9iDKjei by [email protected] |
|
0 likes, 0 repeats |
|
|
|
@codinghorror If communication doesn’t need to be precise, it’s meaningless. |
|
|
|
Post #AxyzWXjHxJJc20JpbM by [email protected] |
|
0 likes, 1 repeats |
|
|
|
@thomasfuchs that depends who you are talking to and on what subject and for wh… |
|
|
|
Post #AxyziZrDryvE6dlhXU by [email protected] |
|
0 likes, 1 repeats |
|
|
|
@thomasfuchs the normal distributions are pretty tight for a lot of tasks these… |
|
|
|
Post #AxyzojdcQ11AB4LEMy by [email protected] |
|
0 likes, 0 repeats |
|
|
|
@codinghorror @thomasfuchs I do this sometimes and I always require receipts: &… |
|
|
|
Post #AxyzojkM0yPCVxUbrs by [email protected] |
|
0 likes, 1 repeats |
|
|
|
@phil_stevens @thomasfuchs yes, the "there ain't no such thing as a fr… |
|
|
|
Post #AxyzzwXGo1tdMdiTSq by [email protected] |
|
0 likes, 0 repeats |
|
|
|
@codinghorror But you’re talking around the issue I’m describing in my OP.T… |
|
|
|
Post #AxyzzwdIRciVfKXHrE by [email protected] |
|
0 likes, 0 repeats |
|
|
|
@thomasfuchs but a LOT of what humans do is INCREDIBLY repetitive. Not all of i… |
|
|
|
Post #AxyzzwjK5DXNy1M6Fc by [email protected] |
|
0 likes, 1 repeats |
|
|
|
@thomasfuchs your argument boils down to "statistics don't do anything… |
|
|
|
Post #Axz06RmnWwly1V547M by [email protected] |
|
0 likes, 1 repeats |
|
|
|
@thomasfuchs now can EVERY ASPECT OF HUMAN LIFE be captured in statistics? Defi… |
|
|
|
Post #Axz0PR9s5Za9ZtvbzU by [email protected] |
|
0 likes, 0 repeats |
|
|
|
@codinghorror I already explained why this is a wrong analogy. ¯\_(ツ)_/¯ |
|
|
|
Post #Axz0PRFtjAP1sakQNs by [email protected] |
|
0 likes, 1 repeats |
|
|
|
@thomasfuchs Agree to disagree. Galaxy brain statistics are quite useful, if yo… |
|
|
|
Post #AxzLAMj6NjzUWj3stU by [email protected] |
|
0 likes, 0 repeats |
|
|
|
@codinghorror @thomasfuchs what about when someone finds your well regarded rep… |
|
|
|
Post #AxzLAMqBxNf6siNXwe by [email protected] |
|
0 likes, 1 repeats |
|
|
|
@miclgael @thomasfuchs well, if the study was published in Science or Nature, I… |
|
|
|
Post #AxzLZE17tr2mksLXQu by [email protected] |
|
0 likes, 0 repeats |
|
|
|
@phil_stevens @codinghorror @thomasfuchs exactly. I'd argue so much work in… |
|
|
|
Post #AxzLZE7VW89F4fKdNY by [email protected] |
|
0 likes, 0 repeats |
|
|
|
@miclgael @phil_stevens @thomasfuchs depends if you're good at asking quest… |
|
|
|
Post #AxzLZEDt8PFhOSJjKC by [email protected] |
|
0 likes, 1 repeats |
|
|
|
@miclgael @phil_stevens @thomasfuchs I mean, I dunno. Cryptocurrency, nfts, VR,… |
|
|
|
Post #AxzLrQ7VdkO5VK0YzI by [email protected] |
|
0 likes, 1 repeats |
|
|
|
@miclgael @phil_stevens @thomasfuchs I'll be happy to take a 10 year long b… |
|
|
|
Post #AxzLxlBKFLxCYxoezw by [email protected] |
|
0 likes, 1 repeats |
|
|
|
@miclgael @phil_stevens @thomasfuchs that'll be after I win the bet with Ca… |
|
|
|
Post #AxzPZDsEpg5I7wRYIq by [email protected] |
|
0 likes, 0 repeats |
|
|
|
@codinghorror @miclgael @phil_stevens @thomasfuchs I would like no more than to… |
|
|
|
Post #AxzPZE1SHPSOaWkufY by [email protected] |
|
0 likes, 1 repeats |
|
|
|
@neurovagrant @miclgael @phil_stevens @thomasfuchs "it's just not the … |
|
|
|
Post #AxzPc1uNmBBfsymino by [email protected] |
|
0 likes, 1 repeats |
|
|
|
@neurovagrant @miclgael @phil_stevens @thomasfuchs also, I blame salesforce. |
|
|
|
Post #AxzQ4Mj7y1AAYs22gy by [email protected] |
|
0 likes, 0 repeats |
|
|
|
@codinghorror @miclgael @phil_stevens @thomasfuchs valid for pretty much everyt… |
|
|
|
Post #AxzQ4MrHThgWy9qYOu by [email protected] |
|
0 likes, 1 repeats |
|
|
|
@neurovagrant @miclgael @phil_stevens @thomasfuchs Death of a Salesman |
|
|
|
Post #Ay0iYyaPlOxF07f864 by [email protected] |
|
0 likes, 0 repeats |
|
|
|
@codinghorror @thomasfuchs this is incorrect. Weather models do understand weat… |
|
|
|
Post #Ay0iYyjHES2lRboCuW by [email protected] |
|
0 likes, 0 repeats |
|
|
|
@cjensen @codinghorror @thomasfuchs For understandable overviews of how weather… |
|
|
|
Post #Ay0iYys8hV8Ht5xHiy by [email protected] |
|
0 likes, 1 repeats |
|
|
|
@jericson @cjensen @thomasfuchs it's a combination. Massive increase in bru… |
|
|
|
Post #Ay2lsaHIXUxJa2gLbc by [email protected] |
|
0 likes, 0 repeats |
|
|
|
@codinghorror @miclgael @thomasfuchs what about the Lancet? |
|
|
|
Post #Ay2lsklnWxIK7X1qXA by [email protected] |
|
0 likes, 0 repeats |
|
|
|
@codinghorror 🔦🌔(couldn't find a blood moon emoji) |
|
|
|
Post #Ay2lt2gQvdhNg6eofg by [email protected] |
|
0 likes, 0 repeats |
|
|
|
@codinghorror instructions unclear. Tried this on my prostitute and he still di… |
|
|
|
Post #Ay2lt2moXunpztducK by [email protected] |
|
0 likes, 0 repeats |
|
|
|
@alice @thomasfuchs I'll just leave this here.. (runs away) https://en.wiki… |
|
|
|
Post #AyAxqnsxjHCDv0tcKe by [email protected] |
|
0 likes, 0 repeats |
|
|
|
@codinghorror @thomasfuchs While that may be true for narrow use cases, most pe… |
|
|
|
Post #AyAxqzMR42J9VbloIa by [email protected] |
|
0 likes, 0 repeats |
|
|
|
@ipaschke @thomasfuchs maybe the "Truth Social" LLM, but who would po… |
