paytonjjones

9 ms·

61.
by paytonjjones 3mo ago
I've treated people with panic disorder who, if there were no cost involved, would probably have called an ambulance monthly for suspected heart attacks. I wonder how countries with universal healthcare coverage deal with the lack of a
62.
by paytonjjones 3mo ago
> doesn't matter how many more nines you add I don't get this argument. People say the same about autonomous driving. But humans also have some number of nines. If you can get it better than humans, that's better!
63.
by paytonjjones 3mo ago
Yep. It will be interesting to see if things move that way or not. AI scraping might get annoying but a lot of sites (eg e-commerce) have obvious incentives to allow AIs to scrape them for free (or even pay AI companies to prioritize them,
64.
by paytonjjones 3mo ago
These do make up a huge % of tool calls, but I don't think these make up a huge % of tool call failures . I see models fail on tool calls that involve API requests to a specific API, internal or cloned Makefile calls, npm run commands
65.
by paytonjjones 3mo ago
Better business model would be some sort of micropayment setup - allow humans, but make scrapers pay a hundredth of a cent for access.
66.
by paytonjjones 3mo ago
> The cap does not apply to Grok, Mr Musk's own AI system Super misleading headline. Please add "external AI" or "non-Grok AI"
67.
by paytonjjones 3mo ago
I wonder if we'll start to see that pattern with every new release. Tool use likely changes rapidly, so the newest, rather than most intelligent, model may always have an edge.
68.
by paytonjjones 3mo ago
Correct, now let's also talk about US government-funded [research, healthcare, education, construction, foreign aid, intelligence, infrastructure, entitlements] Except almost everyone has their pet topic where they'll defend any a
69.
by paytonjjones 3mo ago
That may well be true, but that hardly applies to the current case of taking $325M in bribes
70.
by paytonjjones 3mo ago
You're presenting this with ironic swipes like "bad things like ask for more money", but it's hard to read this description as anything but straightforwardly more efficient. If there are few downsides to centralizing gam
71.
by paytonjjones 3mo ago
Ha, I was literally thinking this but from the other side. "Hmm, 7MB would barely make a dent in the size of the app and allow us to do some of our basic ML without calling the backend" Probably a lot more practical to use this th
72.
by paytonjjones 4mo ago
Your timeline is a bit backwards: they were given trillions of dollars only after they'd demonstrated a few pretty remarkable advancements. The fact that their advancement suggested that pouring more compute would continue working was
73.
by paytonjjones 4mo ago
BERT is a transformer! The unlock happened within transformers, yes, but they were not exactly super new or innovative architectures at that time. The scale was the main innovation that brought us from BERT to today's frontier.
74.
by paytonjjones 4mo ago
Research and experimentation on neural nets has been going on since the 70s (arguably much earlier even), but the lions share of capability changes has all been in the last couple years. Scale was really the unlock; the new pre and post tr
75.
by paytonjjones 4mo ago
Maybe not AGI, but if you look at the differences between, say, GPT-2 and GPT 5.5, it's remarkable how well it works to mostly just throw scale at the problem.
76.
by paytonjjones 4mo ago
I think everyone wants a high trust society but you can't just remove all guardrails and expect that to be the result. The causality goes the other way.
77.
by paytonjjones 4mo ago
I think this can be explained by the dictum "big effects get discovered first" It does present a weird science communication problem. After the first generation, scientists are all focused on "little effects" and don&#x2
78.
by paytonjjones 4mo ago
> wanting to be perceived as a real-world Tony Stark The thing is, he _had_ this. He had all of Reddit worshipping him. He had the press. He had the tech-internet. Then he knowingly blew it all up overnight to ally with Trump. I think th
79.
by paytonjjones 4mo ago
Can you spell out how the grift works? All I'm understanding from your post is that wealthy people are connected to the effort.
80.
by paytonjjones 4mo ago
That sounds horribly toxic and corrosive for a dinner party. It sounds pretty useful for when you're chatting while waiting for the bus and there's someone on drugs there screaming obscenities. Unfortunately the Internet is both.
81.
by paytonjjones 4mo ago
This is a common takeaway from Intro to Psychology but unfortunately it's just...not very true to life (or to psychological science). It's a fault of us psychologists. The most interesting studies are those that are surprising, so
82.
by paytonjjones 4mo ago
That's great messaging, but what exactly are you materially proposing that's different? If you make corporations liable for minors using their product, they're just going to require identity verification to use their product,
83.
by paytonjjones 4mo ago
As is common in social science, it's really challenging to disentangle this kind of thing. But there are ways you can get at it. One clever example discussed in the blog I linked: > We also found five studies that used a similar des
84.
by paytonjjones 4mo ago
No, I don't think it's the same as an unfalsifiable or circular claim. Network level effects are still falsifiable, it's just often really challenging to falsify them at scale.
85.
by paytonjjones 4mo ago
Yes. The research is quite confusing. This is because the strongest version of the argument is not "your child uses social media, and that makes them depressed". The strongest version is more like "when a society mass adopts
86.
by paytonjjones 4mo ago
I think your reply and the parent can both be true, you're just using slightly the same words to describe different things. The parent is talking more about elegant simplicity vs. sprawling, seemingly haphazard complexity, and you&#x27
87.
by paytonjjones 4mo ago
It's not that they are planning the sequence and understanding all the cogs in the machine. It's that they've seen the same situation play out 10 times already with other kids. All they need to do is copy it exactly.
88.
by paytonjjones 4mo ago
I don't think that follows. Say you had a perfectly smooth progressive chain from rocks to spears to guns to nuclear weapons. When it comes to government restrictions, you still have to choose to draw lines somewhere, right?
89.
by paytonjjones 4mo ago
That's always been a relative, rather than absolute statement. Genuine question: if Democrats take power, do you expect them to be more interventionist or less interventionist with respect to AI? Bernie's jockeying leads me to sus
90.
by paytonjjones 4mo ago
The current state of the world, where we have insane and ubiquitous surveillance tech but our packages are nevertheless being constantly stolen (with the thieves "caught" under said surveillance but with no one bothering to enforc

More ›