Lifeboat Foundation: Safeguarding Humanity
Toggle light / dark theme

Blog

Category: innovation – Page 80

I ve been quite impressed so far. And, if they can be improved over night i would love to see it.

With long-term memory, language models could be even more specific – or more personal. MemoryGPT gives a first impression.

Right now, interaction with language models refers to single instances, e.g. in ChatGPT to a single chat. Within that chat, the language model can to some extent take the context of the input into account for new texts and replies.

In the currently most powerful version of GPT-4, this is up to 32,000 tokens – about 50 pages of text. This makes it possible, for example, to chat about the contents of a long paper. To find new solutions, developers can talk to a larger code database. The context window is an important building block for the practical use of large language models, an innovation made possible by Transformer networks.

Continue reading “MemoryGPT is like ChatGPT with long-term memory” | >

Lead author Yurii Victorovich Kovtun, despite being forced to evacuate the Kharkiv Institute of Physics and Technology amid the current Russia-Ukraine war, has continued to work with Kyoto University to create stable plasmas using microwaves.

Getting plasma just right is one of the hurdles to harnessing the massive amounts of energy promised by nuclear fusion.

Plasmas — soups of ions and electrons — must be held at the right density, temperature, and duration for atomic nuclei to fuse together to achieve the desired release of energy.

Read more | >

(0:00) Nick kicks off the show.
(1:24) Under the hood of ChatGPT
(7:53) What is a neural net?
(10:05) Cast.ai — Get a free cloud cost audit with a personal consultation at https://cast.ai/twist.
(11:33) Determining values and weights in a neural net.
(18:28) Vanta — Get $1000 off your SOC 2 at https://vanta.com/twist.
(19:33) Emulating the human brain.
(23:26) Defining computational irreducibility.
(26:14) Emergent behavior and the rules of language.
(31:49) Discovering logic + creating a computational language.
(38:10) Clumio — Start a free backup, or sign up for a demo at https://clumio.com/twist.
(39:38) Wolfram’s ChatGPT plugin.
(43:46) The rapid pace of AI
(58:45) The “Post-Knowledge Work” era.
(1:03:52) The unintended consequences of AI
(1:11:45) Rewarding innovation.
(1:16:12) The possibility of AGI
(1:20:07) Creating a general-purpose robotic system.

Check out Wolfram Research: https://www.wolfram.com/

FOLLOW Stephen: https://twitter.com/stephen_wolfram.
FOLLOW Jason: https://linktr.ee/calacanis.

Thanks to our partners:

Continue reading “Stephen Wolfram on AI’s rapid progress & the “Post-Knowledge Work Era” | E1711” | >

Scientists have achieved a “temporal analogue” to the famous double-slit experiment that could lead to new optical technologies. ABSTRACT breaks down mind-bending scientific research, future tech, new discoveries, and major breakthroughs. Scientists have discovered “unexpected physics” by opening up “slits” in time, a new study reports, achieving a longstanding dream that can help to probe the behavior of light and pioneer advanced optical technologies.

Read more | >

We’re half way there! Join us… and find out more!

Subscribe: https://wmojo.com/unveiled-subscribe.

In this video, Unveiled takes a closer look at arguably the most crucial stage of the entire Kardashev Scale — Type 3.5! This is the in between time, when a galactic power is trying to take over the whole universe… and, if humans ever get there, we could be unrecognisable as a species!

This is Unveiled, giving you incredible answers to extraordinary questions!

Continue reading “What If Humanity Was a Type 3.5 Civilization? | Unveiled” | >

In a quest to expand human settlement to the moon and beyond, NASA and industry partners are developing landing systems to take astronauts from orbit to the lunar surface with its mission Artemis. However, a small but pertinent issue of excessive dust on the surface is posing a considerable challenge for fulfilling such explorations.

To counter this, the agency is seeking ideas from university students for solutions to mitigate the issue related to dusty landings. NASA’s new Human Lander Challenge invites college students to research ways to regulate the plume effect, the process in which a cloud of dust is stirred up by lunar vehicles when rocket engines are used to provide thrust for a smooth landing on unprepared surfaces. NASA had conducted a similar program for students, through its Breakthrough, Innovative, and Game-changing challenge in 2020 for solutions to the same issue.

Read more | >