Related Posts
New here. Henloooo
Bowl Leader
Additional Posts in Consulting
Book recommendations for a vacation?
HMU if interested in a referral to Point B ! Management consulting firm based in Seattle, offices across the country
- Fantastic WLB: 35-40hrs normal, and never more than 50 (and people will check in on you if you do lol)
- Generous time off: ~6-7 weeks is the norm, I’m taking 9 this year
- Flexible remote work: Offices are open, but no formal return to office as of now
- Lots of other random benefits/perks like wellness stipends and fully paid for annual getaways per office for you and a +1
Any 🐠 here from CGI federal!? How is it there?
New to Fishbowl?
Download the Fishbowl app to
unlock all discussions on Fishbowl.
unlock all discussions on Fishbowl.
Understatement. Pentagon labelled them a supply chain risk, which means none of the vendors that do business with the DoD can have anything to do with them six months from now.
Fun fact, all three of the major cloud vendors (AWS, Azure and GCP) do business with the DoD. They not only use Anthropic models, they supply Anthropic with the compute it needs to train its models. Those cloud vendors have made major investment plans into data center capacity in anticipation of Anthropic being one of their biggest customers.
Hegseth is basically threatening to blow up a $350bil dollar American company that has given the DoD significant capabilities (and potentially the entire economy with it), because he doesn't want to agree in writing not to spy on the American public and use LLMs for murderbots. Absolutely insane.
Time will tell, but considering his multi-billion dollar engineering team all came out in favor of Anthropic, he'd be a fool to make that claim without at least a fig leaf in place to support it.
If I had to bet though, this does stink of OpenAI maneuvering behind the scenes to displace Anthropic in the defense and public sector space since they're at significant risk of falling behind them and Google on the commercial side currently.
Chief
I would love to work for Anthropic! It’s already hard enough to get it, now it will be impossible.
Anthropic just took a stance that is not really relevant in this decade. Two decades ago mass surveillance / Snowden revelations were new and alarming. Now with social media and cameras all over, public privacy is a myth. So, they could have soften their stance and added more software guardrails to prevent breaking of laws.
Chief
The legal action is against the supply chain risk designation they got, it’s unprecedented for US companies. Hope they win!
The fact that thy just gave the contract to OpenAI after Altman said they would have the same red lines shows how much of a dumb farce this is.
This will get settled. Team Trump doesn’t want to pop the AI market bubble.
I mean when the DOW asked them point blank if they could use their models to stop an incoming ballistic missiles attack and Anthropics response was “you need to call us an we’ll handle it”, it’s clear you’re not dealing with a party that actually wants to reach a reasonable deal. Imagine if Raytheon or Palantir said the same ridiculous things to the DOW. The response would be the same.
LLMs also aren't designed to calculate ballistic missile trajectories either, so the point stands regardless. If the DoD tried to use these models in ways they're not designed for and people get hurt or killed as a result, Anthropic could be held legally liable, so it makes perfect sense for them to have contractual language that limits that product liability.