
We're an AI research company that builds reliable, interpretable, and steerable AI systems. Our first product is Claude, an AI assistant for tasks at any scale.
We're an AI research company that builds reliable, interpretable, and steerable AI systems. Our first product is Claude, an AI assistant for tasks at any scale.
Launched on April 3rd, 2025
Launched on March 26th, 2025
I love @Cursor. It's enabled me to build (vibe code) so many web apps, sites, extensions, and little things quickly that 1. bring me joy and 2. help me with work or realize personal projects.
However... I'm seeing a TON of movement around @Claude by Anthropic's Claude Code. I haven't personally tried it but it's apparently insane (and can also be expensive?)
I'm curious. Should I switch? What are you currently using? Or do they both have their own use case. I right now like cursor because I can build directly in a GitHub repo or locally and it helps me learn my way around an IDE.
Looking forward to hearing everyone's thoughts!
Product Hunters,
Let's talk about Anthropic's Claude. Everyone praises its focus on safety and responsible AI, which is admirable. But I can't help but wonder: does this intense safety alignment sometimes come at the cost of raw capability or uncensored utility, especially when compared to rivals like GPT-4?
Is "safety" becoming a convenient justification for certain limitations, or is it genuinely paving the way for a more trustworthy, albeit potentially more cautious, AI? What are your thoughts on this balance? Does Claude's "helpful and harmless" sometimes feel... too careful for real-world innovation?
Hit me with your honest opinions.
Anthropic just released two new models for Claude 4: Opus 4 and Sonnet 4. Let's find out how good they are.
What are you using it for?
What s impressed you? What s broken?
What I love most is how it gets the context right almost every time. Whether I'm asking for help with code debugging, content ideas, or analyzing some data, it keeps track of our conversation and gives relevant answers. It feels less robotic and more like having a really smart conversation partner.
Probably it's not appropriate for the simple use case I tested on. Asked: what are the 10 sub niches that can be used for a brand new blog, with no subscribers and no social media yet (of course, it's brand new), that can generate revenue and profit within 60 days. The answer was extremely simple: only listed 5 high level overused topics (niches) like health, finances, crypto, which is literally impossible to make a cent in 60 days. Instead any other (Grok, Gemini, Perplexity) provided right off the bat interesting subtopics (not necessarily is guaranteed will make money within 60 days, but at least they eliminated the overused/overcrowded from the get go.