Stories
Slash Boxes
Comments

SoylentNews is people

SoylentNews is powered by your submissions, so send in your scoop. Only 17 submissions in the queue.
posted by hubie on Friday May 19 2023, @04:08AM   Printer-friendly
from the johnny-5-alive dept.

https://arstechnica.com/information-technology/2023/05/anthropics-claude-ai-can-now-digest-an-entire-book-like-the-great-gatsby-in-seconds/

On Thursday, AI company Anthropic announced it has given its ChatGPT-like Claude AI language model the ability to analyze an entire book's worth of material in under a minute. This new ability comes from expanding Claude's context window to 100,000 tokens, or about 75,000 words.

Like OpenAI's GPT-4, Claude is a large language model (LLM) that works by predicting the next token in a sequence when given a certain input. Tokens are fragments of words used to simplify AI data processing, and a "context window" is similar to short-term memory—how much human-provided input data an LLM can process at once.
[...]
While it may not sound impressive to pick out changes in a text (Microsoft Word can do that, but only if it has two documents to compare), consider that after feeding Claude the text of The Great Gatsby, the AI model can then interactively answer questions about it or analyze its meaning. 100,000 tokens is a big upgrade for LLMs. By comparison, OpenAI's GPT-4 LLM boasts context window lengths of 4,096 tokens (about 3,000 words) when used as part of ChatGPT and 8,192 or 32,768 tokens via the GPT-4 API (which is currently only available via waitlist).
[...]
Notably, Anthropic received a $300 million investment from Google in late 2022, with Google acquiring a 10 percent stake in the firm.

Anthropic says that 100K context windows are available now for users of the Claude API, which is currently restricted by a waitlist.


Original Submission

This discussion was created by hubie (1068) for logged-in users only, but now has been archived. No new comments can be posted.
Display Options Threshold/Breakthrough Mark All as Read Mark All as Unread
The Fine Print: The following comments are owned by whoever posted them. We are not responsible for them in any way.
(1)
  • (Score: 2) by MIRV888 on Friday May 19 2023, @07:33AM

    by MIRV888 (11376) on Friday May 19 2023, @07:33AM (#1306966)

    At least have it read a decent book.

  • (Score: 2) by Opportunist on Friday May 19 2023, @08:09AM (1 child)

    by Opportunist (5545) on Friday May 19 2023, @08:09AM (#1306972)

    Can it produce a book report where it answers the questions that damn teacher is asking that the usual book report crib sheets don't answer?

    • (Score: 2) by gnuman on Friday May 19 2023, @07:36PM

      by gnuman (5013) on Friday May 19 2023, @07:36PM (#1307054)

      Of course it can. But can the students then answer some oral questions about the book that they have apparently read and get provide meaningful answers. Or will they just stand there with silent, gaping mouths like ghostly figures adrift in the eternal green light, their silent cries swallowed by the opulence of their gilded surroundings.

      (Aside, ChatGPT finished the sentence in the spirit of Great Gatsby)

  • (Score: 2) by Rich on Friday May 19 2023, @10:19AM

    by Rich (945) on Friday May 19 2023, @10:19AM (#1306978) Journal

    Not having a $300M investment? Not enjoying a waitlist queue? To a proprietary API?

    These guys did the Gatsby challenge before, they do 64K+ tokens, and their model can freely be downloaded for local inference:

    https://huggingface.co/mosaicml/mpt-7b-storywriter [huggingface.co]

    (Gatsby example down on the model card page)

    I wonder if there's any relation to this Anthropic stuff.

  • (Score: 3, Insightful) by SomeGuy on Friday May 19 2023, @06:06PM

    by SomeGuy (5632) on Friday May 19 2023, @06:06PM (#1307043)

    Claude AI Can Now Digest an Entire Book

    Nomnomnomnom BUUUURRRP!

    While it may not sound impressive to pick out changes in a text

    It isn't. So what. A book is a small amount of data compared to what computers can and do deal with these days. So their shit is highly optimized? Whoopety cowshit, who cares?

    But these days, if a magical "AI" farts, it is somehow news.

(1)