- Refactored.io
- Posts
- Anthropic's 10x AI 🤖
Anthropic's 10x AI 🤖
Plus: AI can crack passwords
Hey folks,
Elon Musk and Apple cofounder Steve Wozniak along with 1,100 prominent researchers signed an open letter calling for 6-month ban on AI. On the other hand, Google’s ex-CEO Eric Schmidt says China might take a lead on AI. Either way, what we need is gov’t regulations to setup safe operating rules.
On to the update:
Anthropic plans for a 10x AI model
AI is getting better at cracking passwords!
Gov’t regulations on AI.
Reading time is 4 minutes. Let’s go!
The Main Thing
You should know
Anthropic plans for 10x AI model
What: Last couple of months have seen disruptive advances in AI. A leaked investor pitch deck from Anthropic, an AI startup backed by Google, revealed what next couple of years might look like.
Similar to ChatGPT, Anthropic currently offers a chatbot named Claude. And now, its working on on creating an updated version, fittingly named Claude-Next.
The new model will be 10x better than today’s version and will need $1 billion over the next 18 months to develop.
VCs are pouring money in AI startups.
Overall, Anthropic is looking to raise $5 billion in next 3 years! Developing AI is neither easy nor cheap!
What's next: 2 key insights from Anthropic's pitch deck -
The AI models could begin to automate large portions of the economy.
The companies that train the best 2025/26 models will be too far ahead for anyone to catch up in subsequent cycles!
Bigger Picture - Although ChatGPT is all the rage right now, Anthropic's Claude has business friendly policies right out of the box. For instance, it follows industry-standard practices for handling enterprise data, comes with APIs for integration and is easy customizable.
An AI 10x better than today would change how we work and create!
Tech Roundup
Stay Informed
PassGAN: New AI tool can crack 51% of common passwords in under a minute, and up to 71% in a day! The AI uses Generative-Adversarial-Network (GAN) to learn the distribution of passwords that have been leaked over last couple of years. Perhaps it’s time stop using passwords and switch to passkey.
Regulations: Gov'ts around the world are starting to build regulations on the use of AI. Italy banned the use of ChatGPT, Germany is considering a similar move.
In the US, California and 17 other states have already introduced bills on AI, while the Federal gov’t is working on a nationwide AI policy.
On the flip side, India stands out as they have opted not to impose any regulations.
Around the Internet
🧠 LinkedIn: offering 100 new courses on AI for free until June!
📊 Graphs: use AI to build graphs from your CSV data.
🔖 Substack: After a public drama with Elon, Substack is launching a twitter-like feature called notes.
🐞 Bug-Bounty: OpenAI launched a bug-bounty program with awards between $200 and $20,000 for every bug.
🛠️ Prompt Engineering: A guide on how to how to use LLMs like ChatGPT and craft prompts to do your tasks.
Misc
How did we do? |