kenjackson

6 ms·

1.
by kenjackson 5d ago
Yes. I forgot about that section. That is probably the one thing that the AI companies can and should do better on.
2.
by kenjackson 5d ago
I didn't know this article was about that issue at all. Yeah, if the issue is properly citing work then yes, OpenAI needs to do that. But the article read like it was tackling a completely different issue.
3.
by kenjackson 5d ago
"how to make sure that, as AI changes the way work is done, we do not lose sight of what that work was meant to achieve in the first place." That sounds like an "us problem", not an AI or OpenAI/Anthropic problem.
4.
by kenjackson 5d ago
Good essay. The author seems to recognize his hypocrisy and why there won't be a lot of sympathy from society. As someone who wrote his first program in 1980, I've actually been really excited about AI. This morning I already wr
5.
by kenjackson 7d ago
I update my phone now when the battery requires two full charges to 100% to last throughout the day. But there's no feature I'm looking for. My current phone is a 13 -- so I guess it lasted five years, which seems crazy long for
6.
by kenjackson 7d ago
Everyone in my family wants one... assuming it works well. We spend a lot of time on our phones, and a lot watching video. This seems like the ideal device for that.
7.
by kenjackson 7d ago
Strongly agree. And as one of the major AI companies, this is extremely tone deaf. If they saw a human (even if assisted) was making great progress on a major problem then you give them space. You don’t swoop in with millions in token spe
8.
by kenjackson 7d ago
But you’re seeing huge effects in three years, not a generation. The report notes “The change in mean scores that can be explained merely by the changing composition of the student body, however, never exceeds a few points.” So for Portugal
9.
by kenjackson 8d ago
Not that quickly though.
10.
by kenjackson 8d ago
What a great story! I'm not sure why I felt teary eyed reading it, but I was. If you're here in the comments but haven't read it -- it is worth the read.
11.
by kenjackson 10d ago
The problem with school children is most books that are assigned have no volta for them. Part of it is that modern storytelling has a much quicker volta through tv, movies, YouTube and now shorts. It’s hard to justify to a kid a 90 minute
12.
by kenjackson 12d ago
I actually started by saying "reasoning" is a strong statement because as you note, it's really not reasoning. It's about intuition. And yes, if you've seen it before then intuition doesn't play a role. But
13.
by kenjackson 12d ago
Reasoning is a strong statement here. But it is fair to say that it usually is not intuitive to us. An example is if I gave you a huge sheet of thin paper (huge so that folding isn’t an issue) - how many times could you fold it in half unti
14.
by kenjackson 12d ago
Can definitely write college level essays and have for a while. The jobs is that when LLMs first started getting popular, but aren’t quite common professors were that some of the worst students in class started writing the best essays. Now
15.
by kenjackson 14d ago
They don't all suck equally. Here's the order, from best to worst. Amodei Google SamA Zuck Elon
16.
by kenjackson 16d ago
I agree. LLMs write really well, but probably too consistently in terms of style. That said, I did recently ask AI to do an editing pass over some writing I did, and I was surprised that it missed a tense mismatch that I had (not within a
17.
by kenjackson 16d ago
> honestly curious question: do you expect this to remain true? I It's already the case that it's becoming not true. For example see this post from Lin Yang: https://x.com/lyang36/status/2092092709251
18.
by kenjackson 19d ago
If Cyclospora is listed on the label then sure. I don't use the builtin speaker on my laptop -- can I get refund on that too?
19.
by kenjackson 20d ago
There also seems to be some state logic as well. The card knows when I’m in the middle of a trip and doesn’t update the balance until a future tap when it can determine the extent of the trip (presumably). Very cool that it’s all done local
20.
by kenjackson 23d ago
How so? I’ve been working on DNNs for over a decade. Not sure it helps me in any almost non-trivial way when it comes to using them.
21.
by kenjackson 23d ago
“It's best not to take advice on direction of careers from anyone. It's better to find and work on things that interest you, and then take advice from people that are amazing in that specific field.” What bucket should I put this
22.
by kenjackson 23d ago
Except knowing how LLMs work don’t actually provide much understanding for using them. People don’t use LLMs the way we’ve built on most other tons or platforms. It’s more learning Unity hoping to be a better gamer.
23.
by kenjackson 1mo ago
What I still struggle with is the huge percentage of people that support the behavior. Seeing what the Tate brothers did and seeing the sheer number of people who cheered on the crimes and exploitation they did was shocking. The fact that
24.
Snapchat to no longer spotlight AI generated content(newsroom.snap.com)
2 points by kenjackson 2mo ago|0 comments
25.
by kenjackson 2mo ago
A property of lack of intelligence though is stopping too soon. The wisdom to differentiate is probably rare at the extremes.
26.
by kenjackson 2mo ago
Yes. Or even to go a step further -- once we exclude all the things that aren't hard to predict, we're left with the things that are.
27.
by kenjackson 2mo ago
He's not wrong though. In the 80s I'd watch video shows and over the course of a week I'd probably see some videos 10 times. And it wasn't background filler - I'd actually be sitting in a chair/couch and watch
28.
by kenjackson 2mo ago
To be clear -- this wasn't meant to be a GPT vs Claude post. This is more that frontier models continue to incrementally get better. No human steering for this level of math is crazy. Go back ten years and ask anyone in the field if
29.
GPT 5.6 Solves all IMO 2026 questions with no human steering [pdf](github.com)
7 points by kenjackson 2mo ago|3 comments
30.
by kenjackson 2mo ago
GPT 5.6 Pro apparently solved all the IMO 2026 problems (competition happened July 15/16 in China). It was done with no human steering and on first attempt. I wasn't involved at all, just ran across this and thought it was impre

More ›