Jessica - if you are reading the comments, I have to say - Founders at Work changed my career trajectory. I read it fresh out of college in 2008. I told a buddy to read it and it also changed his trajectory.
Guess what, after years of meandering (YC, Series A, big tech) I still come back to the moment I first discovered your book.
Also, please tell PG, I knew about your book before I knew what YC was :-)
Other comments already mentioned multiple services (from OpenAI to Cleanspeak). I want to provide a high level clarification from experience.
Moderation is a vast topic - there are different services that focus on different areas: such as, text, images, CSAM, etc. Traditionally you treat each problem area differently.
Within each area, you, as an operator, need to define the level of sensitivity for the category of offense (policies).
Some policies seem more clear cut (eg image: porn) while others seem more difficult to define precisely (eg text: bullying or child grooming).
In my experience, text moderation is more complex and presents a lot of risks.
There are different approaches for text moderation.
Keyword based matching services like Cleanspeak, TwoHat, etc. are baseline level useful but limiting because assessing a keyword requires context. A word can be miscategorized and results in false positive or false negative with this approach, which may impact your operation at scale; or UX if a platform requires more of a real-time experience.
LLM is theoretically well suited for taking context into account for text moderation; however they are also pricier and may require furthering fine tuning or self-hosting for cost savings.
CSAM as a problem area presents the highest risks though may be more clear cut. There are dedicated image services and regulatory bodies that focus on this area (for automating reporting to local law enforcement).
Finally, EU (DSA) also requires social media companies adhere to self report on moderation actions. EU also requires companies to provide pathways for users to own and delete their data (GDPR).
Edit: FIXED typos; ADDED a note on CSAM and DSA & GDPR
Case in point, under Case Study > Reconnaissance > OSINT, these two paragraphs follow one another - same content but different wording.
> The first step in any hardware hacking project is research. I started by Googling the router model number, "ASUS RT-N12 D1", and came across an article about a similar model, the ASUS RT-N12+ B1. The article mentioned that the device had an open UART interface allowing unauthenticated root access. However, it provided no exact details on how to exploit this or where the UART interface might be located. Could my router model have the same vulnerability?
> In the first step I googled the model number for my router "ASUS RT N12 D1" and I came accross this article. It shows that a similar model the "ASUS RT N12+ B1" appears to have an open UART interface, which gives unauthenticated root access. It does not show how to exacltly abuse this or any details where to find the UART interface. Let's see if our router model may have the same vulnerability!
What’s your observations on non-big pharma - ie small late stage clinical trials startup with a single drug. Would they not vastly benefit from joining a standard protocol?
> Initially, I was shocked at how slowly everything moved! I was used to working really quickly when producing and engineering albums. Suddenly it was like the brakes were on and often it was difficult to get the momentum going. Eventually, I adapted to the Floyd pace. One of the great things about working with this band is that you are allowed time to be creative, to pursue an idea even if it takes some time. The Floyd had a production deal to make their records and the record label never heard anything until it was done. The record was made purely and only by the people in the studio.
The creative freedom without commercial intervention - this is very cool. I can almost hear it in The Wall - how grand and elongated the songs are.
I got around to building v3 of the Airtable GTD tracker / dashboard template to share. Feel free to make a copy and get started; link in blog post.
Airtable syncs with my Google Calendar, and I’ve configured it to send me a daily digest. It’s typically in my inbox by 5:30AM Pacific Time, so I get to see what I have due that day.
Signals Music Studio channel on YouTube has a 30 min long video [0] on music theory map (knowledge graph) in VR. It’s not particularly in depth but spells out hope each concept connects to others.
It feels like the approach OP is taking won’t be able to take this short cut into account.
I too am interested in linking up good bike trails. Mostly for the east bay gravel systems. Today, I save GPX or geojson from routes I find on Strava and import into a map client (CalTopo). It’s a okay solution but my problem is in finding more alternative routes.
Can someone explain the leading and lagging indicator in this paragraph?
> OKRs, given they typically focus on change are usually leading indicators focused on shorter time windows and KPIs, which represent the health of what the organisation seeks to sustain, are usually lagging indicators focused on longer timeframes.
Guess what, after years of meandering (YC, Series A, big tech) I still come back to the moment I first discovered your book.
Also, please tell PG, I knew about your book before I knew what YC was :-)