@mattshumer_

–

09 April 2024 16h00

Remember, OpenAI, Anthropic and DeepMind are training 10x bigger models than Opus

Using Claude Opus with Console Interface Features

By

–

09 April 2024 15h57

Opus will likely be noticeably better for your use case. Don’t use the http://
Claude.ai UI, use the one at http://
console.anthropic.com. Less polished, but it allows you to edit generations, delete messages, set system prompts, etc.

Essential Python Function for AI Engineers

By

–

09 April 2024 5h37

Here is one of the most powerful and reusable functions you'll need as an AI engineer: “`
def remove_first_line(test_string): if test_string.startswith("Here") and test_string.split("n")[0].strip().endswith(":"): return re.sub(r'^.*n', '', test_string, count=1)

Claude 3 Opus outperforms GPT-4 for task delegation

By

–

09 April 2024 5h14

Claude 3 Opus feels like the first model I can really hand off tasks to without writing a really detailed spec. GPT-4 was always quite capable but even for simple tasks, you really needed to poke and prod it. Not so with Claude.

Evaluation Loss Analysis Across Multiple Training Runs

By

–

09 April 2024 4h56

eval loss of many training runs

NYC Early-Stage Startups Outpace SF Innovation Landscape

By

–

09 April 2024 1h06

The early-stage startups that have really excited me lately have all been in NYC. Some public, many still in stealth. Feels like people are actually trying unique, interesting, and useful things here, whereas in SF it's a lot of the same overlapping boring ideas.