The CFA Institute just published a primer on cryptoassets for investment professionals. It's nice and concise and covers the following:

- The Basics - How Crypto Works and Why It Matters
- Understanding the Crypto Landscape
- Crypto as an Investment Opportunity
- Crypto in a Portfolio Setting
- Key Considerations and Risks for Investors

⚡️ Alibaba announced it will split into six main businesses. In time, each will be able to pursue fundraising and IPOs.

Alibaba becomes six Little Babas

The real revolution in AI is still to come - former Meta-AI engineering manager and VP of AI/ML at Cruise Hussein Mehanna

If you think GPT-4 and similar systems approach human-level intelligence, ask yourself why any teenager can learn to drive a car in about 20 hours of practice and yet we still don't have Level-5 self-driving cars. This is despite having enormous amounts of training data from expert drivers and vehicles equipped with sensors with superhuman capabilities.

Game changer. You can now run GPT locally on your macbook with GPT4All, a new 7B LLM based on LLaMa.

It's completely open source: demo, data and code to train an assistant-style large language model with ~800k GPT-3.5-Turbo Generations based on LLaMa.

Independent implementation of LLaMA that is fully open source under the Apache 2.0 license.

With a rush to adopt generative AI, use of open-source models is booming.

As some of the biggest companies use models from online repositories, security researchers say they need to be cautious because some may contain dangerous security flaws, bias or not work as promised.

Of course, OpenAI's ChatGPT also had a recent security bug.

Welcome to the new security challenges in the age of more rapid AI adoption. One company, Robust Intelligence, is unveiling a tool to scan models for security, effectiveness and bias.

Finally an open-source version of DeepMind's Flamingo

A framework for training and evaluating Large Multimodal Models (LMMs) capable of processing images and text.

This model is trained with StabilityAI support and compute. Comes with all sorts of goodies, like a multi-model dataset, which is needed for the future.

Blog.
Demo.

A new theory: Spatial Computing. The brain computes by using brain waves to control where info is expressed in its networks.

Working memory control dynamics follow principles of spatial computing.

“We don’t even understand the brain of a worm” - Christof Koch of the Allen Institute thinks we have a long way to go before we truly understand the human brain.

Here are 5 fundamental questions about the brain that still need answers

CZ tweeted that Binance would be converting its $1 billion in BUSD for the Industry Recovery Initiative to "native crypto," boasting "transparency."

Instead, the funds ended up back in a corporate exchange wallet, where they will remain.

March of 2023 will go down as one of the most revolutionary months in history.

AI developments released this month changed the world forever.

Here's the rundown of the biggest events that happened.

Survey of Large Language Models

China's third largest state-owned insurance institution Pacific Insurance Investment Management Hong Kong Branch and Waterdrop Capital launched a compliant blockchain venture capital fund and POS token income enhancement fund in Hong Kong.

Impressive. LLMs can self-improve without additional training data, reinforcement learning, or human intervention.

1. Generate an initial output
2. Analyze its own output and provides feedback.
3. Improves its output based on its own feedback.

Beauty is in the eye of the medial prefrontal cortex (mPFC).

Caltech researchers used machine learning and fMRI to understand how the brain quickly decides whether or not we enjoy a piece of art, much the same way the brain decides whether we enjoy food, giving new meaning to "taste in art."

It increasingly looks like the EV market might play out not like the smartphone market but like the Android market: a tidal wave of new competitors and a war for market share.

Microsoft published a 155-pages long paper about the capabilities and limitations of an early GPT-4.

Quote from the abstract: GPT-4 could reasonably be viewed as an early (yet still incomplete) version of an AGI.

MUST READ: 59 car makers are actively building in Web3 and the Metaverse.

So check out this report from the team at Labs3, which includes:

- Loads of charts and data on which car companies are most active
- Industry outlook for use of Blockchain, NFTs and the Metaverse
- A VERY cool use case matrix covering the different applications and players

Brilliantly AI announced SightBot is a ChatGPT-powered chatbot you can chat with about biomedical research topics and receive responses with PubMed citations!

Italy’s decision to ban ChatGPT is likely to be just the start of OpenAI’s potential troubles in Europe.

Germany is contemplating following Italy’s lead. The U.K. data privacy regulator has warned companies that if they aren’t careful their use of generative AI models could land them in hot water too.

Some legal scholars are wondering if any large foundation model could ever really comply with Europe’s strict data privacy law, GDPR—and whether that means GDPR itself might need to be changed.

What's the benefit of tokenising everything on a Blockchain? $16 TRILLION! (Or 10% of Global GDP)...

That's the value of assets that Boston Consulting Group (BCG) forecast will be tokenised using Blockchain technology by 2030.

That's a very big number, and we're currently at a pretty low base today. So where and why are we going to see this forecast explosion in tokenisation of real-world and digital assets in the next 7-years?

The team at Rejolut (no, not the neo-bank) have pulled together some helpful case studies so you can catch-up quickly. They've included examples covering:

- Precious Metals
- Real Estate
- Artwork
- Agro Commodities
- Securities
- Medical and Pharmaceutical Products
- Finance
- Gaming...

...and they have put together a helpful forecast of how these sectors are likely to continue their path to adoption and innovation; including some of the ongoing challenges (e.g. practice-based evidence, trust, regulation and infrastructure).