Chapter 47.1: Thursday Want to see what happens when someone famous calls you at 2am with a favor you can't check? Oh, and what it costs to tell a friend you gave her name away... "Then stop answering your phone. I mean that. Thank you for picking up."
common failure mode is picking the tech that gives you the fastest early ramp, riding that steep curve, feeling like you're winning, projecting that, you know, steep slope into the future and feeling like the sky is the limit, and then hitting the plateau, and discovering that the technology path that you picked actually flattens out way before the performance that is required by your product.
And as a result of identifying that person in a lineup, that person then replaces what they actually saw. So our memory can be contaminated just like any form of evidence could be contaminated.
if you ask a person when was Abraham Lincoln born and they don't know the date. They could sit there, they could think about it for a week, if they if they don't have access to Wikipedia or something, they're not going to be able to do better answering that question if they thought about it for a week compared to 5 seconds. Same with the model.
And I believe the best product CEOs are in the minutia of the details. you know, all the stories about Steve Jobs. I was so impressed when I heard that he insisted on picking out the carpeting in the conference rooms, you know, and and I tried to have that same mentality of micromanaging the really small pixel level details that matter.
And it's the first time I learned the difference between activities and impact. activity is you being an all-star answering a bunch of calls, but the ticket queue for the team is still high.
I mean, you could argue that we we're already past peak truth on the on the on the internet, right? And now there's there's just more and more garbage every every day. It gets harder and harder to suss out the stuff that you do want to include in the training set in order to actually make something more intelligent.
Uh but whenever we do a systematic study, um any given problem, an AI tool has a success rate of maybe 1 or 2%. Uh it's just that it's just that they can apply at scale and and you just pick the winners, it looks great.
And I realized it was the first one on the list and I thought, huh, I wonder if people are just picking the first one. So then then we randomized the list so everyone saw a different order of the list and now all the items were picked equally
The Thursday Murder Club by Richard Osman is a delightful whodunnit set at a retirement home in the English countryside. Great characters. Fun twists. An endearing sense of humor. Nothing fancy or experimental, just a pleasure to read (and what more can you ask from a book, really?). If, like me, you enjoy mysteries, you'll be in good hands with Osman.
You don't call uninformed. You call knowing as much as there is to know and then you start picking the brains of people who might know more and have not said.
So picking the right question is the hardest part of science and making the right hypothesis. And that's what today's systems definitely they can't do. So I often say it's harder to come up with a conjecture, a really good conjecture than it is to solve it.
It's a lot easier for someone to engage with an argument if they generated the key steps themselves by answering my questions - if imposed by me, it sparks contrarianism and defensiveness
It’s totally fine to have a turtle. Everyone has a turtle. You can’t build a theory without a turtle. It depends on the problem you want to describe. Actually, the reason I can’t get behind Stephen’s ontology is I don’t know what question he’s trying to answer. Without a question to answer, I don’t understand why you’re building a theory of reality.
That’s the only way how you deal with propaganda, because propaganda is not necessarily something that is an outright lie. It can be just one factor that’s taken out of the context and is blown out of proportion. And that is good enough.
I would argue having a business leader run these institutions and then having a board that has, itself, diverse viewpoints, and by the way, permanently structured to have diverse viewpoints is a much better way to run a university than picking an academic that the faculty supports.
At a higher level, the real event was a catastrophic failure of the industry's strategy of relentlessly shouting as loud as they can "Hey, get off our case, we're busy saving lives here!"
Text is a much more powerful mode for model outputs. A model that can generate images can only be used for image generation, whereas a model that can generate text can be used for many tasks: summarization, translation, reasoning, question answering, etc.
Just dumping the code on GitHub is not open source. Open source is a culture. Open source means that your issues are not all one year old, stale issues. Open source means developing in public.
This, as was noted on Twitter, is a recipe for bad studies: it encourages the cherry-picking of results that went your way and the hiding of those that didn’t.
If you’ve got a small genome, the chances of you picking up the right bit of DNA from the environment is much higher than if you’ve got a genome of 20,000 genes. To do that, you’ve effectively got to be picking up DNA all the time, all day long and nothing else, and you’re still going to get the wrong DNA. You’ve got to pick up large chunks, and in the end, you’ve got to line them up, you’re forced into sex, to coin a phrase.
Such restrictions are sub-optimal for sentence-level tasks, and could be very harmful when applying fine-tuning based approaches to token-level tasks such as question answering, where it is crucial to incorporate context from both directions.
Such restrictions are sub-optimal for sentence-level tasks, and could be very harmful when applying fine-tuning based approaches to token-level tasks such as question answering, where it is crucial to incorporate context from both directions.
Such restrictions are sub-optimal for sentence-level tasks, and could be very harmful when applying fine-tuning based approaches to token-level tasks such as question answering, where it is crucial to incorporate context from both directions.
Such restrictions are sub-optimal for sentence-level tasks, and could be very harmful when applying fine-tuning based approaches to token-level tasks such as question answering, where it is crucial to incorporate context from both directions.
In the real world, picking up a new language takes a few weeks of effort and after 6 to 12 months nobody will ever notice you haven’t been doing that one for your entire career.
A korrent is a belief a person has stated in their own words: one
sentence stating the claim, backed by a quote and a source, kept at
korrents.com.
Under a name here, the quoted block is what they actually said.
The korrent beneath it is the claim those words support, in
korrents' wording — tap it to see the record, its source, and who
else holds it.
Nobody here wrote their own korrents. They are compiled from public
statements, and a person can change their mind, which is recorded too.
About the English under a post
Some people here publish in a language other than English. Where they
do, this site shows a machine translation beneath the post, in
this typeface — the site's own, not theirs.
The post itself is never changed, moved or hidden: what is set in the
serif above is exactly what the person published, and it is what to
quote them on. A translation can be wrong in ways that matter,
especially about tone.
Only the post's own words are translated. A quoted post, a linked
article and a belief on korrents.com
are left in their original language.