ttul

42 ms·

661.
by ttul 3y ago
I think the title is misleading. This is an MIT PR piece.
662.
by ttul 3y ago
The nuance is that their technique allows you to slide the context window along without introducing a jagged discontinuity in generation. At least that’s my feel for it.
663.
by ttul 3y ago
No worries. My initial response was horrible; trying to do better.
664.
by ttul 3y ago
Appreciate the skepticism; it keeps us on our toes. Let's cut to the chase: Domain Lockdown: We added this to our Cloudflare Workers integration. It mandates a DNS TXT record to authorize a Worker to send emails from a specific domain.
665.
by ttul 3y ago
It’s basically a way of more efficiently making use of memory transfers during the calculation of the attention blocks in a transformer. You transfer a block at a time, increasing inference throughout because less time is spent overall fetc
666.
by ttul 3y ago
Appreciate the skepticism; it keeps us on our toes. Let's cut to the chase: Domain Lockdown: We added this to our Cloudflare Workers integration. It mandates a DNS TXT record to authorize a Worker to send emails from a specific domain.
667.
by ttul 3y ago
All Workers API calls to MailChannels must send email from a domain that has a Domain Lockdown record authorizing the Worker by its CF-Worker header. Docs are here: https://support.mailchannels.com/hc/en-us/article
668.
by ttul 3y ago
I interviewed JGC in June about the rich technical history of Cloudflare. In this long-format discussion, John opens up about many of the intriguing technical things Cloudflare did as they scaled. I hope this is interesting to many of you h
669.
Long-format interview with Cloudflare's CTO [video](youtube.com)
6 points by ttul 3y ago|1 comments
670.
Canada's broadcast regulator fights spam and phishing in 2023 [video](youtube.com)
2 points by ttul 3y ago|1 comments
671.
by ttul 3y ago
In June, I interviewed Steven Harroun of the CRTC (Canada's sort-of equivalent of the FTC), exploring the complexities of Canada's Anti-Spam Legislation, enforcement against cyber abuse, and the challenges AI brings to the table.
672.
DMARC's Impact and New Threats from Language Models Like ChatGPT [video](youtube.com)
4 points by ttul 3y ago|1 comments
673.
by ttul 3y ago
Mike Jones of Fortra Agari Discusses the Evolution of Anti-Phishing, Role of DMARC, and the New Threats Posed by Language Models like ChatGPT. A deep dive into email authentication and its future challenges.
674.
Spamhaus will protect the internet from AI-driven spam [video](youtube.com)
3 points by ttul 3y ago|1 comments
675.
by ttul 3y ago
Here is my interview with Spamhaus talking about what they are doing to protect the internet from bad stuff. We talk about the history of Spamhaus, the challenges of generative AI in the email space, and how Spamhaus is evolving.
676.
by ttul 3y ago
… I should have carved out a space for cold prospecting. I’m using GPT-3.5 to detect it in my own inbox because bulk spam filters don’t seem to have the requisite power to deal with such highly customized messages at scale.
677.
by ttul 3y ago
It’s a nuance, but when I refer to “spam” I am incorporating the term “phishing.”
678.
by ttul 3y ago
I disagree. The “open” part implies there is no control at all on what email can be relayed. That is not at all the case here. Spammers try their best to send spam through MailChannels all day long, through compromised WordPress sites, hack
679.
by ttul 3y ago
SendGrid send roughly 8% of all internet email. At their scale, controlling abuse is extremely difficult. Gmail originates a huge amount of spam and phishing. The bigger you are, the more spammers will work to target you and the harder you
680.
by ttul 3y ago
What about mailing lists? How about mail forwarding?
681.
by ttul 3y ago
I've worked in anti-spam since 2003. I am skeptical that his claims about ARC work at scale. Google does not have a rules-based spam filtering system. It is entirely driven by a massive machine learning model that adjusts its weights i
682.
by ttul 3y ago
Thought experiment: Would an email sending service that sends email for 42% of the domains globally that send through a transactional service [1] possibly exist on today's internet if it _wasn't_ incredibly good at blocking spam?
683.
by ttul 3y ago
Nobody in the email world considers ARC to be a bulletproof way to bypass spam filtering at a major receiver. The DEFCON presenter was ill-equipped to make that determination.
684.
by ttul 3y ago
https://community.cloudflare.com/t/this-week-mailchannels-wi...
685.
by ttul 3y ago
Why is it not valid for someone to send email from another domain? I’m asking seriously. That’s not an invalid thing to do. The RFC explicitly advises that flexibility in the envelope address is a necessary feature of internet email. It may
686.
by ttul 3y ago
Absolutely. For reference, here is a TXT lookup for a random large corporation: % host -t txt bosch.com bosch.com descriptive text "b5r8gr465ydgqlvwlwy6x7dshccwg37c" bosch.com descriptive text "mindmanager-verification=5eacd5
687.
by ttul 3y ago
You're pointing out the inherent challenge of domain authentication. Large domain owners like Google lock their domains down by publishing restrictive SPF records that do not authorize anyone other than their own IP space. For everyone
688.
by ttul 3y ago
Except it's not an open relay. MailChannels would not exist on the internet if it didn't aggressively control spam and phishing. The DEFCON talk did not prove the existence of a gaping hole; it showed something that has existed si
689.
by ttul 3y ago
The next release of DMARC will likely include an option to exclude SPF from DMARC verification. Google's team is pushing for this on the IETF DMARC mailing list: https://mailarchive.ietf.org/arch/msg/dmarc&#x2
690.
by ttul 3y ago
Note that since June 2023, it is impossible to send email from Workers through MailChannels unless you publish a _mailchannels record in your DNS.

More ›