Anthropic

We’re rolling out access to Claude.ai to more people around the world.

Starting today, users in 95 countries can talk to Claude and get help with their professional or day-to-day tasks. You can find the list of supported countries here: https://lnkd.in/dN6_qrpT

Since launching in July, millions of users have leveraged Claude’s expansive memory, 100K token context window and file upload feature. Claude has helped them analyze data, improve their writing and even talk to books and research papers. Now users in all supported countries can access both our free experience and Claude Pro to boost their productivity and get more done.

We’re working hard to responsibly expand availability over the coming months—and will have more to share soon.

New Anthropic research: How university students use Claude.

We ran a privacy-preserving analysis of a million education-related conversations to produce our first Education Report.

Our findings show students primarily use Claude to create and improve educational content (39.3%) and provide technical explanations (33.5%).

Computer Science dominates AI usage, representing 38.6% of Claude conversations despite making up only 5.4% of US degrees.

We created a taxonomy for how students interact with AI, and tracked how common those interaction styles were across different subjects. Science and Math saw more Collaborative Problem Solving; Education saw more Direct Output Creation.

Our taxonomy reveals distinct interaction patterns: more Collaborative Problem Solving in Science/Math and Direct Output Creation in Education fields.

We also wondered, are students using AI to cheat?

It’s hard to know whether any specific category is cheating or legitimate study, but we did notice some concerning patterns: students straightforwardly asking for answers, or asking Claude to rewrite an essay to avoid plagiarism detection.

Students mostly used AI to create and analyze, according to our analysis using Bloom’s Taxonomy of educational goals.

Whether cheating or not, this raises important questions about how to stop students offloading important cognitive tasks to AI. AI can empower learning in remarkable ways. But we’re under no illusions that all the changes it’s bringing to education are positive.

Our research will help educators and policymakers ensure that AI deepens, rather than undermines, learning.

Read the full report: https://lnkd.in/eABqfPta

We’re working with universities to help realize the benefits of AI. Last week we announced Claude for Education, with a new Learning Mode that helps develop students’ critical thinking skills.

It's available for all Pro users with an .edu email: https://lnkd.in/eP5wGyn7

The Claude iOS app has arrived. The power of frontier intelligence is now in your back pocket.

Chat with Claude whenever, and wherever, inspiration strikes. Claude on iOS syncs with your chat history, lets you ask questions about photos, and handles tasks on the go — powered by our industry-leading Claude 3 models.

Download now on the App Store: https://lnkd.in/gNhrqyXm

We are pleased to announce Claude 2, our new model! Claude 2 has improved performance, longer responses, and can be accessed via our API as well as a new public-facing beta website, claude.ai - currently available in the US and UK.

We have heard from our users that Claude is easy to converse with, is less likely to produce harmful outputs, and has a longer memory. We have made significant improvements from our previous models on coding, math, and reasoning. For example, our latest model scored a 71.2%, up from 56.0%, on the Codex HumanEval, a Python coding test. And Claude 2 scored 76.5% on the multiple choice section of the Bar exam, up from 73%.

Although no model is immune to jailbreaks, we’ve used a variety of techniques, as well as extensive red-teaming, to make Claude 2 more harmless and harder to prompt to produce offensive or dangerous output. We have an internal red-teaming evaluation that scores our models on a large representative set of harmful prompts, using an automated test, while we also regularly check the results manually. In this evaluation, Claude 2 was 2x better at giving harmless responses compared to Claude 1.3.

Thousands of businesses are working with Claude, and we are pleased to offer them access to Claude 2 via API at the same price as Claude 1.3! Partners like Jasper and Sourcegraph are working with us to power new AI applications that prioritize safety and reliability.

Read more about Claude 2 and our new open beta chat experience here: https://lnkd.in/gfSHAT7s

We're expanding Claude's Research capabilities with an advanced mode that searches the web, your Google Workspace, and now your Integrations too.

Claude now automatically determines when to search and how deeply to investigate.

With Research mode toggled on, Claude researches for up to 45 minutes across hundreds of sources (including connected apps) before delivering a report, complete with citations.

Available today in beta for Max, Team, and Enterprise plans, and soon to the Pro plan.

Read more: https://lnkd.in/eTE6DxXi