Kyle Maclean

[Kyle Maclean @AnalyticsKyle]

@KikuchiBeliever @cecidrais I agree that it is not "evil". It's just a waste of limited time. I think they ask to make themselves feel clever and be able to argue for whichever prospect they wanted to draft in the first place.

[Kyle Maclean @AnalyticsKyle]

@KikuchiBeliever @cecidrais Is there any evidence that any of this is predictive of their ability to play hockey in the NHL? It seems like any answer might be good or bad, and all this is does is confirm whatever your previous thinking was.

[Kyle Maclean @AnalyticsKyle]

@zenahitz You're right that it was sudden but ot was entirely organic. Chatgpt had essentiallly no marketing whatsoever other than users telling each other about it.

[Kyle Maclean @AnalyticsKyle]

@TheLincoln Would "lmao" be categorized as "bluster"? Do you have data to provide evidence for your point?

[Kyle Maclean @AnalyticsKyle]

@TheLincoln Economic work is some of the highest quality empirical work out there.

[Kyle Maclean @AnalyticsKyle]

@Jackbmeyer Students could already teach themselves almost any introductory university course through textbooks and videos. I don't think access to the content was the binding constraint.

[Kyle Maclean @AnalyticsKyle]

@erinhengel Maybe. All I'm saying is that there are other hypotheses for why he helped. :)

[Kevin A. Bryan @Afinetheorem]

One reason I started teaching my "progress" class is the vibes vs reality gap. Coming into the greatest decade in human history: basically solved metabolic syndrome, deaths from autos, huge progress in abundant green energy and cancer, going back to the moon, (and yeah, AI) 1/2

[Kyle Maclean @AnalyticsKyle]

@erinhengel Another hypothesis is that it an instance of "an enemy of my enemy is friend", and her actual innocence is irrelevant to that.

[Kyle Maclean @AnalyticsKyle]

@KelseyTuoc Wait was it well known that Edelman was her counsel? I didn't know that.

[Kyle Maclean @AnalyticsKyle]

Note the "multiple outs" setup here. Rewritten, the prediction is that Gemini 3.5 pro will either be better or worse than GPT 5.5 Pro.

[Kyle Maclean @AnalyticsKyle]

This is a very helpful paper for an ongoing and important conversation. And like any paper, feels obvious once done.

[Peter John Lambert @pj_lambert]

Is GenAI causing the relative decline in early-career hiring? Our latest research finds that these effects may be conflated with another important driver: the rise of WFH arrangements (1/N)

[Kyle Maclean @AnalyticsKyle]

Apparently Machu Picchu was built in the 15th century. So...granted I'm an idiot, but am I the only one who just assumed it was built like 3000 years ago?

[Kyle Maclean @AnalyticsKyle]

I love this as an example of very cheap technology (i.e., recorded music) not supplanting the human taste/preference for nevertheless watching other people perform a task live. There is opportunity for both!

[Dudes Posting Their W’s @DudespostingWs]

Dude activity: watch Lord of the Rings on a giant screen while a live orchestra performs the entire score in sync with the movie

[Kyle Maclean @AnalyticsKyle]

@1929stockcrash You know that if he defined milk tea without asking, he would then be accused of mansplaining.

[QC @QiaochuYuan]

at this point it is completely untenable to believe anything along the lines of “AI can only spit out an average of the training data.” that was already only a very rough way of understanding older models pre-reasoning, it was already obsoleted by o1 which released in 2024, and now it should be obviously and conclusively dead even if you haven’t been paying close attention. recursive self-improvement has barely even started and we are already here. even with the recent erdos problem solves you could argue that those were cherrypicked out of a large database for being neglected by humans. that cope is no longer available

now generalize the lesson: all other arguments that there is some essential human activity forever beyond the reach of AI are also cope, these are technical problems and the will and money and talent exists and is being deployed to solve them. artificial superintelligence is not a fairy tale. assume it’s coming and plan accordingly

[Kyle Maclean @AnalyticsKyle]

@Afinetheorem You quietly just said something really important.

[Kyle Maclean @AnalyticsKyle]

@ontheceiling87 @TheStalwart I think they'll want evidence of human authenticity on some dimensions/contexts and not others. For instance, perhaps you want your therapist to be human and giving you attention, but you care less if the restaurant reservation line is.

[Kyle Maclean @AnalyticsKyle]

@SouthernWintrs @StarkConor @PhilipDBunn A certain % of our job is to teach. You may wish it was split among positions but that isn't the case yet.

[Kyle Maclean @AnalyticsKyle]

@PhilipDBunn FWIW, it is extremely rude (and completely irrelevant) to quote someone's citations at them to shut down an argument. You were more gracious here than I would have been.