nikcub

5 ms·

31.
by nikcub 4mo ago
Google are the only one of the big three who can tick the boxes on being multimodal, price / performance and having Apple-level of compute available
32.
by nikcub 4mo ago
> "a well written agents.md is very good for the agent" while even a mildly bad agents.md can be _very_ bad for the agent. they rot very quickly which is why human curation is essential. same with memory - a lot of the self-lea
33.
by nikcub 4mo ago
Bright Data is available as a product on AWS Marketplace https://aws.amazon.com/marketplace/seller-profile?id=bf9b432...
34.
by nikcub 4mo ago
Same - have accounts with 3 different services and have never been KYC'd even with heavy usage.
35.
The Smart TV in Your LivingRoom Is a Node in the AIScraping Economy(blog.includesecurity.com)
235 points by nikcub 4mo ago|105 comments
36.
by nikcub 4mo ago
> which actually lead to a #1 spot on openrouter usage that was only because it was free
37.
by nikcub 4mo ago
It's becoming apparent that it requires more tokens to secure code than it does to write it May even be an order of magnitude more
38.
by nikcub 4mo ago
there are 150M+ of them and you'll be taking out a lot of human users with it modern blocking is behaviour / heuristic based
39.
by nikcub 4mo ago
Cloudflare are more likely to be undercounting bots - they don't really pick up many of the modern browser-driven bots and crawlers.
40.
by nikcub 4mo ago
these old network security techniques don't really work anymore. the common bots are at known IP ranges, the problem bots are all on datacenter + residential proxies.
41.
by nikcub 4mo ago
and an iPhone 17e for $599
42.
by nikcub 4mo ago
I believe the urgent deprecation timeline here may be related to ai labs using offline licensed Office in agents as part of workflows and Office integration. Microsoft wants _each_ agent instance to be a separate license[0] There was always
43.
by nikcub 4mo ago
Same price on a token basis, but usually steadily decreasing on a task basis
44.
by nikcub 4mo ago
The syntax is the easier part - most programming tasks require the reasoning and understanding of a large world model to solve problems. Fine tuning a 'lean and smart' model works really well for discrete, repeatable high volume t
45.
by nikcub 5mo ago
the CVE + patch data has been built into models for a few generations now. I actually thought the bug bounty companies were well positioned here, but they've been overtaken. Mythos is a better hacker than we ever were
46.
by nikcub 5mo ago
Finding the neeedle is easier when you remove the haystack Or providing a map with a direction There is a long history of high-value private vulns being rediscovered from scant details
47.
by nikcub 5mo ago
There has been a lot of cynicism around mythos, that it's just the usual public models without guardrails, etc. etc. but this: > 1,752 of those high- or critical-rated vulnerabilities have now been carefully assessed by one of six i
48.
by nikcub 5mo ago
Anthropic this quarter will have revenue of $10.9B, up from $4.8B last quarter[0]. They're paying SpaceX $1.25B per month for compute[1] - which is more than what SpaceX earn on space. SpaceX spent about $30-40B in capex on Colossus 1
49.
by nikcub 5mo ago
This just incentivizes market for bio-mules, which already exists with world[0] - where prices stay low because it was rolled out to low-income countries. Then there's the platform game theory. If you adopt you add friction which reduc
50.
by nikcub 5mo ago
This is the conflict at the center of running a hosting company - make it easy to signup and you get a lot of new users but also a lot of abuse. Implement anti-abuse measures and you will hit some loud false positives (this may be the case
51.
by nikcub 5mo ago
I'm trying to think of a standard this doesn't apply to
52.
by nikcub 5mo ago
Because they later founded a for-profit to raise the capital required in the AI talent and compute race.
53.
by nikcub 5mo ago
both realtime markets where multi-AZ is hard?
54.
by nikcub 5mo ago
oh they have too! came out a couple of months ago.
55.
by nikcub 5mo ago
They're so tied to crypto that i'm surprised they haven't been tempted to diversify into other asset classes, or even yolo into prediction markets like robinhood did. It would be slop, but the market would love it
56.
by nikcub 5mo ago
can't believe Bret's bio doesn't mention friendfeed!
57.
by nikcub 5mo ago
the AI replies itt are cringe
58.
by nikcub 5mo ago
* real programmers write assembly, not FORTRAN * real programmers manage memory, it's a craft * real programmers don't drag and drop * real programmers don't use intellisense * real programmers don't need stack overflow
59.
by nikcub 5mo ago
> If a tree falls in the forest... I hope this is a pun on the content management system used to publish OP. It's forester[0], written in OCaml and parses TeX-like .tree files into semantic XML which uses browser XSLT to render the
60.
by nikcub 5mo ago
the most cited is terminal bench 2.0, but its also plagued by cheating accusations and benchmaxxing. somewhat remarkably, claude code ranks last for Opus 4.6 - which may say something about cc, or say something about the benchmark [0] http

More ›