Stat Health brings in $5.1M for its in-ear wearable to track cerebral blood flow

J2 Ventures, BonAngels Venture Partners and a diverse group of angel investors backed the company through the seed funding.

Stat Health also received grant funding from the U.S. Air Force.

Stat Health designed its wearable to help understand symptoms like dizziness, brain fog, headaches, fainting and fatigue upon standing.

These common symptoms can indicate illnesses like long COVID and postural orthostatic tachycardia syndrome (POTS). They also may signal myalgic encephalomyelitis/chronic fatigue syndrome (ME/CFS), and other orthostatic syndromes.
According to Stat Health, reduced blood flow to the brain upon standing causes the symptoms for these illnesses.

The company clinically tested its offering at Johns Hopkins, and it was peer-reviewed in the March 2023 issue of the Journal of the American College of Cardiology (JACC). Stat Health said it demonstrated the ability to predict fainting minutes before it happens.

Carve out a few hours to learn streamlit.io
Powerful for rapid prototyping, interactive visualization.
It's a hammer and you'll start seeing a lot of nails.

Are we at the beginning of a new era of small models? Here is newest LLM trained fully at Microsoft Research

phi-1 achieves 51% on HumanEval w. only 1.3B parameters & 7B tokens training dataset.

Any other >50% HumanEval model is >1000x bigger (e.g., WizardCoder from last week is 10x in model size and 100x in dataset size).

Corporates Using NFTs - Are They More Than a Passing Fad - Deloitte

NFTs, and the distributed blockchain networks behind them, represent a breakthrough in digital rights management as well as digital representations of assets. Here are just some of the other possible breakthroughs NFTs could drive: 

1 Building a bridge between the digital and physical worlds to authenticate and provide evidence of a transfer

2 Democratising ownership of digital collectibles—for example, the creation of new ways to monetise art, photographs, music, intellectual property (IP), and more

3 Selling digital items—homes, high-end sneakers, streetwear, and more—for use with avatars in gaming and online worlds

4 Developing “super wallets” that allow an NFT owner to keep a verified record of all licenses and rights, along with product warranties, event tickets, access passes for secure locations for work or leisure, and more 

5 Securing the ticketing industry against fraud, providing a percentage of secondary sales revenue to performers or venues, and creating unique keepsakes

6 Extending and monetising brands in new ways for both existing and new customer bases

7 Offering utility services—for instance, serving as a VIP card granting access to a secret concert, membership in an exclusive community, or special discounts on products.

CVPR 2023 announced the Best Paper Awards! It's the world's most prominent computer vision conference, with citation impact just below Nature & Science.

- Best Paper 1: VisProg uses GPT to generate executable code that parses an image and does effective visual reasoning, even though GPT itself is blind. Similar principle as Voyager (executable code to play Minecraft).

VisProg: prior.allenai.org/projects/vispr…
Paper: arxiv.org/abs/2211.11559
Another paper with similar high-level idea is called ViperGPT, worth checking out: arxiv.org/abs/2303.08128

- Best Paper 2: Unified Autonomous Driving (UniAD), a comprehensive framework that incorporates full-stack driving tasks in one network.

Planning-oriented Autonomous Driving: opendrivelab.github.io/UniAD/
Paper: arxiv.org/abs/2212.10156

- Best Paper Honorable Mention: DynIBaR, a new state-of-the-art on synthesizing novel views from a monocular video of a complex dynamic scene.

DynIBaR, Neural Dynamic Image-Based Rendering: dynibar.github.io
Paper: arxiv.org/abs/2211.11082

- Best Student Paper: a new 3D point cloud registration technique that finds the optimal pose to align a pair of point clouds.

3D Registration with Maximal Cliques: arxiv.org/abs/2305.10854

- Best Student Paper Honorable Mention: a diffusion model that can be customized to a particular subject with only 3-5 example images.

DreamBooth: dreambooth.github.io
Paper: arxiv.org/abs/2208.12242

Did you know 1/3 of the world uses 'tonal' languages? For example, the Mandarin word "Ma" may have 4 different meanings based on intonation.

Current english-based speech BCI decoding models could hardly detect tonal differences because they are produced by subtle movements of the larynx.

Published this week, a large consortium in Shanghai came together to overcome this shortcoming.

They also published their ECoG dataset.

Many African, Asian and South American nations are now a step closer to speech restoration.

EY have created a playbook for DAOs - a new way to launch and run companies in Web3 (and Web2.5)

DAO treasuries now have around $25Bn in funding available to build new businesses, or fund growth of their platforms and products.

This report covers everything you need to know to understand the world of DAOs, how they create value, and what it takes to launch one. Including:

- History of DAOs (1991 to today)
- Statistics on the growth of DAOs and DAO members
- Case studies of Web3 and Web2 participation in DAOs
- Key characteristics of a DAO
- Classifications of DAOs (pooling funds, spending time, shared interest)
- DAO value drivers
- Critical success factors for DAOs

Wow! Inflection-1 is a new best-in-class LLM powering Pi, outperforming GPT-3.5, Llama and PALM-540B on major benchmarks commonly used for comparing LLMs.

$1 billion AI venture fund offers elusive Nvidia chips to win deals

Nat Friedman and Daniel Gross, a pair of founders turned startup investors, set the venture capital world abuzz last week by announcing a novel tactic to win over
founders: offering them access to a huge number of free servers equipped with the most advanced chips for training machine-learning models.

It turns out the duo also quietly raised an investment fund with more than $1 billion in assets to invest in artificial intelligence and infrastructure startups.

Rival venture capitalists say they are closely watching Friedman, the former CEO of Microsoft-owned GitHub, and Gross, a former investment partner at startup accelerator Y Combinator, to see whether their server-chip offer to founders gives them an edge in landing AI startup deals.

The chips have been in extraordinarily short supply as demand for AI has heated up.

The move by Friedman and Gross comes as funding for young AI startups has ballooned, especially for those developing LLM that can automate tasks related to writing and software code.

Friedman and Gross said they recently acquired 2,512 Nvidia H100 server chips, which could be worth $100 million, that they said they would give to founders in exchange for equity—in addition to giving the startups capital from the duo’s venture fund.

The fund, C2 Investments, which started raising capital in the past two years, was managing assets valued at $1.1 billion as of March, according to the regulatory filing. It isn’t clear whether that figure includes the chips and servers.

Other VC firms have debated allocating part of their funds to purchasing chips to help them win deals. Spark Capital, a VC firm whose AI bets include OpenAI rival Anthropic, has discussed dedicating $50 million of the firm’s fund to providing founders with access to AI hardware, according to a person with direct knowledge of the matter.

OpenAI turned to Microsoft for access to graphics processing unit servers in exchange for equity in the startup and a share of OpenAI’s future profits, plus access to its technology.

Precision Neuroscience gets brainy - human brainy. First ever. Great coverage for the technology and the team

Founded in 2021 by a co-founder of Neuralink, Elon Musk's brain-computer interface startup, Precision Neuroscience is an industry competitor working to help patients with paralysis operate digital devices by decoding their neural signals.

Several companies like Synchron, Paradromics and Blackrock Neurotech have also created devices with this capability. 

On Saturday a hackathon dedicated to AI agents organized by AGI House took place

One of the highlights was Jim Fan, senior AI scientist at Nvidia.

Here are his slides on open-ended agent learning:

- Voyager: GPT-4 lifelong learning agent
- MineDojo: open-source Minecraft AI framework
- MineCLIP: RLHF for embodied agents.

Video

New paper from the Meta AI Lab: “Dimensionality and ramping: Signatures of sentence integration in the dynamics of brains and deep language models”

They investigated how the brain represents the semantic composition of multiple words into a sentence using simultaneous MEG and intracranial EEG.

They introduced to language processing a neuroscience concept that was previously found useful to analyze neural responses to simpler stimuli: the intrinsic dimensionality of a neural manifold.

They predicted that as successive words are composed, the vector space spanned by a set of sentences should become exponentially larger, thus yielding a roughly linear increase in dimensionality. They verified this in simulations of large language models.

And they found the predicted increase in intrinsic dimensionality in both sEEG and MEG signals ! It is particularly large when the sentence is meaningful, relative to control stimuli made of letter strings or of pseudowords (Jabberwocky).

They further used decoding to dissect 3 types of processes that distinguish normal and jabberwocky sentences: phasic lexical access, sustained integration (ramping), and end-of-sentence wrap-up.

They found that different regions exhibit each pattern (red = phasic lexical, blue = ramping integration, green = end-of-sentence wrap up).

🔥The future of the оperating room

Animares and MagicLeap powered AR in the operating theater.

AR and VR are transforming healthcare by offering immersive experiences and revolutionizing medical training, patient care, and diagnostics, leading to remarkable advancements in the field.

⚡️ Biotech backed by Fosun Group says it has begun the first mid-stage human trial of a drug entirely "discovered and designed" by AI.

The latest report by the World Economic Forum entitled "Top 10 Emerging Technologies of 2023" outlines the technologies poised to positively impact society in the next three to five years.

Among them, the Metaverse for mentalhealth: "Excess screen time and social media can decrease psychological well-being, but they can also enhance well-being when used responsibly. Screen time spent building connections in shared virtualspaces might help combat the growing mental health crisis as opposed to contributing to it."

The NHS is rolling out its first AI program as a medical imaging device to save clinicians time and improve radiotherapy outcomes.

The head and the prostate are two locations where particular care is needed during radiotherapy to avoid accidental damage to surrounding tissue, which could lead to long-term neurological damage or incontinence.

And as each tumour has a unique shape, this requires up to 2 hours of consultants' time to carefully plan each patient's therapy.

AI has been shown to accelerate this process considerably by suggesting a first pass of the contours for treatment.

Clinicians will always check the outputs, but in 2/3 of cases they approve its suggestion without needing any corrections.

Following successful trials and safety checks in Cambridge, the cloud-based program is now going to be offered at cost price to hospitals across England.

Passionate about solving Loneliness & Social Connection with cutting-edge AI? Amorai are hiring a Head of Operations

Founded by former Tinder CEO, Renate Nyborg, and backed by Andrew Ng’s AI Fund, Amorai’s mission is to help one billion people master the skill of social connection.

As Amorai prepare to launch first product - an AI-powered coaching app tackling Gen Z loneliness, and relationship health - they are seeking a dynamic Head of Operations to lead and optimize day-to-day activities.

🔥Shanghai released a plan to build out its culture and tourism metaverse projects into an industry generating an annual revenue of 50 billion yuan (US$6.9 billion) by the end of 2025.

Hong Kong rating agency HKVAC announced its virtual asset index, in addition to BTC ETH stablecoins, including:

WBTC, BTCB, stETH, BCH, LTC, BNB, MATIC, ADA, ATOM, FIL, NEAR, ALGO, ICP, XRP, DOGE, DOT, WTRX, SHIB, TRX, LINK, LEO, XMR, XLM, OKB, LDO, HBAR, VET, QNT, FTM, EOS.

This institution is supported by Hong Kong legislators Johnny Ng, Huobi, KuCoin and others.

However, the listed tokens include platform tokens, privacy tokens, and a large number of tokens listed as securities by the US SEC, so the Hong Kong community believes that the reference value is not high.

The Hong Kong SFC requires retail-tradable tokens to be included in at least two indices. This index is non-governmental, and its judging criteria are simple, that is, the top 30 by market capitalization.

Muscle sensor company sees early promise in a pig

Phantom Neuro, a small startup working on a muscle sensor to allow better control of prosthetic limbs, has tested its device on a pig and found that could decode the pig's movements in real time.

The device would communicate the amputated limb's movements to the person's phone, and after that, to the prosthetic.

The company team will continue working on making the sensor communication more seamless, and eventually testing it on sophisticated robotic limbs.

In the meantime, Phantom Neuro has built out its board with Thomas Oxley, CEO of brain-computer interface company Synchron, and Joshua Duyan, former Chief Strategy Officer of CTRL labs. The company has raised around $9 million, according to AlphaSense.

Google has killed its Project Iris AR smart glasses. It’s now focused on building an AR software platform for 3rd party device makers (and not just Samsung).