Oobit partners with LG to bring interoperable metaverse platform to TVs
Oobit operates a metaverse platform, has partnered with LG Electronics to bring interoperable virtual worlds to LG TVs. Through this partnership, Oobit and LG will bring immersive games and experiences together and make it easy for consumers to interact in the metaverse.
“We’ve been working on the hardware, networking, and software layers to create the metaverse for almost a decade – it just wasn’t called the metaverse back then,” said Pooya Koosha, CTO and Oorbit cofounder, in a statement. “Our proprietary technology is the connective tissue that links virtual worlds together and makes it easy for developers and brands to bring their experiences into the metaverse. Scaling our technology for millions of LG TV customers is the next step in making the metaverse accessible for all.”
Using cloud streaming technology, Oorbit users will be able to enjoy super high-fidelity interconnected virtual worlds and experiences including virtual concerts on Elynxir from Pixelynx and AI generative multiplayer games in Auxworld from Auxuman.
Oobit operates a metaverse platform, has partnered with LG Electronics to bring interoperable virtual worlds to LG TVs. Through this partnership, Oobit and LG will bring immersive games and experiences together and make it easy for consumers to interact in the metaverse.
“We’ve been working on the hardware, networking, and software layers to create the metaverse for almost a decade – it just wasn’t called the metaverse back then,” said Pooya Koosha, CTO and Oorbit cofounder, in a statement. “Our proprietary technology is the connective tissue that links virtual worlds together and makes it easy for developers and brands to bring their experiences into the metaverse. Scaling our technology for millions of LG TV customers is the next step in making the metaverse accessible for all.”
Using cloud streaming technology, Oorbit users will be able to enjoy super high-fidelity interconnected virtual worlds and experiences including virtual concerts on Elynxir from Pixelynx and AI generative multiplayer games in Auxworld from Auxuman.
YouTube
Oorbit x LG x Elynxir - CES2023
👍3
John May of John Deere says that we are losing our farmland even as the world population heads from 8 billion today toward 10 billion soon. Tech has to provide food security to the world. CES2023
2 years ago today, Open AI introduced DALL-E. Aditya Ramesh, DALL-E's inventor and DALL-E 2's co-inventor, says he is "surprised" by its impact.
“Research like the development of GANs in 2014 and DeepMind’s WaveNet in 2016 were already starting to show how AI models could generate new images and audio from scratch.”
“Research like the development of GANs in 2014 and DeepMind’s WaveNet in 2016 were already starting to show how AI models could generate new images and audio from scratch.”
VentureBeat
Two years after DALL-E debut, its inventor is “surprised” by impact
Two years ago today, OpenAI announced the debut of DALL-E. Inventor Aditya Ramesh says is he "surprised" by the model's massive impact.
We are now in 2008 (2022). Big financial crash. The first iPhone (ChatGPT) was presented just last year. The first functional Android (1.0) phone will be presented in October later this year. The next 3-5 generations will revolutionize literally everything.
Next 3-5 generations of ChatGPT and its future competitors will revolutionize literally everything in the same way smartphones did in the past 15 years. People who haven’t realized this today will do in the next 3 years.
Value creation will be decent but nothing compared to the crazy disruptive stuff that will come in the following years.
Cost to build on this minimal, costs scale with revenue.
Next 3-5 generations of ChatGPT and its future competitors will revolutionize literally everything in the same way smartphones did in the past 15 years. People who haven’t realized this today will do in the next 3 years.
Value creation will be decent but nothing compared to the crazy disruptive stuff that will come in the following years.
Cost to build on this minimal, costs scale with revenue.
🔥3
New work from MetaAI: HyperReel. Looks like VR will get a new killer app
Capture videos with multiple cameras set up at different angles → Run HyperReel → You can now step *into* the dynamic scene and freely walk around
Essentially a high-res 4D experience replay.
HyperReel enables "6 Degree-of-Freedom video". It means a VR player can change their head position (3 DoF) and orientation (3 DoF), and the view will be synthesized accordingly. HyperReel is based on the NeRF technology (Neural Radiance Fields).
The biggest strength of HyperReel over prior works is the memory and computational efficiency, both crucial to portable VR headsets. It runs 18 frames-per-second at megapixel resolution on an NVIDIA RTX 3090, using only vanilla PyTorch.
Website: hyperreel.github.io
Paper: arxiv.org/abs/2301.02238
Code and models are open-source!!!
https://github.com/facebookresearch/hyperreel
Capture videos with multiple cameras set up at different angles → Run HyperReel → You can now step *into* the dynamic scene and freely walk around
Essentially a high-res 4D experience replay.
HyperReel enables "6 Degree-of-Freedom video". It means a VR player can change their head position (3 DoF) and orientation (3 DoF), and the view will be synthesized accordingly. HyperReel is based on the NeRF technology (Neural Radiance Fields).
The biggest strength of HyperReel over prior works is the memory and computational efficiency, both crucial to portable VR headsets. It runs 18 frames-per-second at megapixel resolution on an NVIDIA RTX 3090, using only vanilla PyTorch.
Website: hyperreel.github.io
Paper: arxiv.org/abs/2301.02238
Code and models are open-source!!!
https://github.com/facebookresearch/hyperreel
GitHub
GitHub - facebookresearch/hyperreel: Code release for HyperReel: High-Fidelity 6-DoF Video with Ray-Conditioned Sampling
Code release for HyperReel: High-Fidelity 6-DoF Video with Ray-Conditioned Sampling - facebookresearch/hyperreel
👍3
The current climate in AI has so many parallels to 2021 web3 it's making me uncomfortable. Narratives based on zero data are accepted as self-evident. Everyone is expecting as a sure thing "civilization-altering" impact (& 100x returns on investment) in the next 2-3 years.
But there's a bull case and bear case. The bull case is way way more conservative than what the median person on my TL considers as completely self-evident. And the actual outcome we'll see is statistically likely to lie in between, somewhat closer to the bear case.
The bull case is that generative AI becomes a widespread UX paradigm for interacting with most tech products (note: this has nothing to do with AGI, which is a pipe dream). Near-future iterations of current AI models become our interface to the world's information.
The bear case is the continuation of the GPT-3 trajectory, which is that LLMs only find limited commercial success in SEO, marketing, and copywriting niches, while image generation (much more successful) peaks as a XB/y industry circa 2024. LLMs will have been a complete bubble.
So far there is *far* more evidence towards the bear case, and hardly any towards the bull case. *But* we're still very far from peak LLM performance at this time -- these models will improve tremendously in the next few years, both in output and in cost.
For this reason the actual outcome we'll see is somewhere between the two scenarios. "AI as our universal interface to information" is a thing that will definitely happen in the future (it was always going to), but it won't quite happen with this generation of the tech.
Crucially, any sufficiently successful scenario has its own returns-defeating mechanism built-in: commoditization. *If* LLMs are capable of generating outsized economic returns, the tech will get commoditized. It will become a feature in a bunch of products, built with OSS.
But there's a bull case and bear case. The bull case is way way more conservative than what the median person on my TL considers as completely self-evident. And the actual outcome we'll see is statistically likely to lie in between, somewhat closer to the bear case.
The bull case is that generative AI becomes a widespread UX paradigm for interacting with most tech products (note: this has nothing to do with AGI, which is a pipe dream). Near-future iterations of current AI models become our interface to the world's information.
The bear case is the continuation of the GPT-3 trajectory, which is that LLMs only find limited commercial success in SEO, marketing, and copywriting niches, while image generation (much more successful) peaks as a XB/y industry circa 2024. LLMs will have been a complete bubble.
So far there is *far* more evidence towards the bear case, and hardly any towards the bull case. *But* we're still very far from peak LLM performance at this time -- these models will improve tremendously in the next few years, both in output and in cost.
For this reason the actual outcome we'll see is somewhere between the two scenarios. "AI as our universal interface to information" is a thing that will definitely happen in the future (it was always going to), but it won't quite happen with this generation of the tech.
Crucially, any sufficiently successful scenario has its own returns-defeating mechanism built-in: commoditization. *If* LLMs are capable of generating outsized economic returns, the tech will get commoditized. It will become a feature in a bunch of products, built with OSS.
👍5
Apple is planning to debut the Reality Pro mixed-reality headset this spring, ahead of WWDC, and ship it later in the year. The push to launch — along with the economy — will lead to an otherwise muted 2023 for Apple.
Apple has started seeding its headset to a small group of third-party developers to begin working on apps for the device and xrOS. It will talk up developer tools at WWDC in June.
Apple has started seeding its headset to a small group of third-party developers to begin working on apps for the device and xrOS. It will talk up developer tools at WWDC in June.
Bloomberg
Apple Will Talk Up Its Mixed-Reality Headset in 2023 But Not Much Else
2023 is set to be the year of Apple’s mixed-reality headset and not much else. Also: The company hikes battery-replacement costs, plans a retail augmented-reality experience and suffers more delays with its new fintech products.
The Hong Kong Financial Secretary stated that the Hong Kong government will soon issue tokenized green bonds for subscription by institutional investors. The government expects to provide the right proportion of regulation for the market to unlock the potential of Web 3.0.
As AI-assisted programming eats up more of the hours spent, it's becoming clear that the "hard" part of the job is now knowing what question to ask or what task to initiate.
This is probably true, or will soon be true, well beyond programming.
This is probably true, or will soon be true, well beyond programming.
6 OpenAI Rivals Google and Microsoft Are Watching
Microsoft’s big bet on OpenAI, whose ChatGPT software can understand and generate conversational text, is starting to look like a stroke of genius.
Microsoft plans to use OpenAI’s tech to improve results in Bing searches and help Word and Outlook customers automatically generate documents and emails using simple prompts.
But OpenAI is far from the only startup developing software that can understand language and that larger tech companies such as Google, Amazon, or Meta Platforms may want to license, partner with or acquire. In the last few years, top researchers from OpenAI, Alphabet’s DeepMind and Google, which pioneered the machine-learning techniques used in ChatGPT, have left those companies to launch or join six startups that compete with OpenAI.
1. Adept AI Labs
Founded: 2021
What it does: Adept developed a browser extension that can automate tasks when given instructions in normal language. For example, given a prompt to find apartment listings in a city within a certain price range, the AI system can surf the web on its own and create a list.
Equity funding: $66 million
Key investors: Greylock Partners and Addition Last valuation: Undisclosed
Employees: 20
2. AI21 Labs
Founded: 2017
What it does: AI21 Labs has three products: AI21 Studio, which allows developers to customize its models for specific purposes; Wordtune, an automated copywriting tool; and Wordtune Read, a program that summarizes documents.
Equity funding: $128 million
Key investors: Walden Catalyst, Ahren Innovation Capital
3. Anthropic
Founded: 2021
What it does: Anthropic develops general-purpose large-language models, with a particular
focus on ensuring its systems are free of negative traits like bias and toxicity. It is an active
publisher of research on how to deploy machine-learning models in a safe way.
Equity funding: $704 million
Key investors: Former FTX CEO Sam Bankman-Fried, Skype founding engineer Jaan Taallinn Last valuation: $4 billion in April 2022, according to PitchBook
Employees: 80
4. Character
Founded: 2021
What it does: Character’s platform lets users create and interact with chatbots that can assume many roles, such as Draco Malfoy or Sigmund Freud. The chatbots can also generate text-based role-playing games that users can customize.
Equity funding: Unknown
Key investors: Angel investor Elad Gil, Gmail creator Paul Buchheit, former GitHub CEO Nat
Friedman, and Ron Conway’s SV Angel and A Capital Last valuation: No valuation in seed round Employees: 18
5. Cohere
Founded: 2019
What it does: Cohere offers an application programming interface that allows developers to add language understanding and chatbot-related functions to their applications. The models can be used for a variety of purposes, but Cohere focuses on summarizing documents, copywriting and search.
Equity funding: $170 million
Key investors: Index Ventures, Tiger Global Management
Last valuation: Between $800 million and $900 million in February 2022, according to a person familiar with the matter
Employees: 175
6. Inflection
Founded: 2022
Location: Palo Alto, Calif.
What it does: Inflection’s co-founders have said little publicly about the company and declined to be interviewed for this article. But its website leaves no doubt about its ambitions in language AI: “Recent advances in artificial intelligence promise to fundamentally redefine human-machine interaction. We will soon have the ability to relay our thoughts and ideas to computers using the same natural, conversational language we use to communicate with people.”
Equity funding: $225 million
Microsoft’s big bet on OpenAI, whose ChatGPT software can understand and generate conversational text, is starting to look like a stroke of genius.
Microsoft plans to use OpenAI’s tech to improve results in Bing searches and help Word and Outlook customers automatically generate documents and emails using simple prompts.
But OpenAI is far from the only startup developing software that can understand language and that larger tech companies such as Google, Amazon, or Meta Platforms may want to license, partner with or acquire. In the last few years, top researchers from OpenAI, Alphabet’s DeepMind and Google, which pioneered the machine-learning techniques used in ChatGPT, have left those companies to launch or join six startups that compete with OpenAI.
1. Adept AI Labs
Founded: 2021
What it does: Adept developed a browser extension that can automate tasks when given instructions in normal language. For example, given a prompt to find apartment listings in a city within a certain price range, the AI system can surf the web on its own and create a list.
Equity funding: $66 million
Key investors: Greylock Partners and Addition Last valuation: Undisclosed
Employees: 20
2. AI21 Labs
Founded: 2017
What it does: AI21 Labs has three products: AI21 Studio, which allows developers to customize its models for specific purposes; Wordtune, an automated copywriting tool; and Wordtune Read, a program that summarizes documents.
Equity funding: $128 million
Key investors: Walden Catalyst, Ahren Innovation Capital
3. Anthropic
Founded: 2021
What it does: Anthropic develops general-purpose large-language models, with a particular
focus on ensuring its systems are free of negative traits like bias and toxicity. It is an active
publisher of research on how to deploy machine-learning models in a safe way.
Equity funding: $704 million
Key investors: Former FTX CEO Sam Bankman-Fried, Skype founding engineer Jaan Taallinn Last valuation: $4 billion in April 2022, according to PitchBook
Employees: 80
4. Character
Founded: 2021
What it does: Character’s platform lets users create and interact with chatbots that can assume many roles, such as Draco Malfoy or Sigmund Freud. The chatbots can also generate text-based role-playing games that users can customize.
Equity funding: Unknown
Key investors: Angel investor Elad Gil, Gmail creator Paul Buchheit, former GitHub CEO Nat
Friedman, and Ron Conway’s SV Angel and A Capital Last valuation: No valuation in seed round Employees: 18
5. Cohere
Founded: 2019
What it does: Cohere offers an application programming interface that allows developers to add language understanding and chatbot-related functions to their applications. The models can be used for a variety of purposes, but Cohere focuses on summarizing documents, copywriting and search.
Equity funding: $170 million
Key investors: Index Ventures, Tiger Global Management
Last valuation: Between $800 million and $900 million in February 2022, according to a person familiar with the matter
Employees: 175
6. Inflection
Founded: 2022
Location: Palo Alto, Calif.
What it does: Inflection’s co-founders have said little publicly about the company and declined to be interviewed for this article. But its website leaves no doubt about its ambitions in language AI: “Recent advances in artificial intelligence promise to fundamentally redefine human-machine interaction. We will soon have the ability to relay our thoughts and ideas to computers using the same natural, conversational language we use to communicate with people.”
Equity funding: $225 million
The Information
Six OpenAI Rivals Google and Microsoft Are Watching
Microsoft’s big bet on OpenAI, whose ChatGPT software can understand and generate conversational text, is starting to look like a stroke of genius. Microsoft plans to use OpenAI’s tech to improve results in Bing searches and help Word and Outlook customers…
$15.3B raised in 2022 digital health venture funding
2022’s $15.3B raised across 572 deals—far lower than 2021's $29.3B and barely surpassing 2020’s $14.7B—signals the tail end of a macro funding cycle that started back in 2019 and peaked with the COVID-19-era investment boom.
What’s unclear is whether we’ve reached the end of this cycle, or if more low funding quarters are on the horizon. With recession concerns looming, there’s a chance that 2023 could be digital health’s lowest venture funding year since 2019.
But downhill runs aren’t necessarily all doom and gloom. Read on for lessons learned from the current market, which companies continued to pull in cash despite tough conditions, and where investors might place their bets in the year ahead.
2022’s $15.3B raised across 572 deals—far lower than 2021's $29.3B and barely surpassing 2020’s $14.7B—signals the tail end of a macro funding cycle that started back in 2019 and peaked with the COVID-19-era investment boom.
What’s unclear is whether we’ve reached the end of this cycle, or if more low funding quarters are on the horizon. With recession concerns looming, there’s a chance that 2023 could be digital health’s lowest venture funding year since 2019.
But downhill runs aren’t necessarily all doom and gloom. Read on for lessons learned from the current market, which companies continued to pull in cash despite tough conditions, and where investors might place their bets in the year ahead.
hfe_booklet_2014_01_27.pdf
18.4 MB
🔥Human in space research – lessons learned and future directions
Ethereum developers eye February public testnet for Shanghai upgrade — link
The Block
Ethereum developers eye February public testnet for Shanghai upgrade
The Shanghai upgrade of the Ethereum network in March will focus exclusively on ether (ETH) withdrawals.
Coinbase announced another 950 layoffs today. In June 2022, Coinbase laid off 1,100 people, accounting for 18% of the total number of employees.
Coinbase said it will shut down several projects with a low probability of success, and will provide at least 14 weeks of basic salary compensation, health insurance and other benefits to the fired employees.
Coinbase said it will shut down several projects with a low probability of success, and will provide at least 14 weeks of basic salary compensation, health insurance and other benefits to the fired employees.
Coinbase
A message from CEO and Co-Founder, Brian Armstrong, to Coinbase employees - Blog
Nigeria and Kenya led Africa startup funding in 2022, according to venture capital tracker Africa TBD. Nigerian fintech Flutterwave and Kenya’s B2B retail platform Wasoko each raised over $100m for the top spots on the leaderboard, reports
Playstream raises $2M for AI-generated livestream content
Playstream has been working on advanced and innovative AI-based solutions for the past two years to help content creators grow by discovering new ways to monetize and gain audiences.
The product is a unique multidisciplinary AI solution, personalized and adaptable to the creator and their communities. Playstream’s solution scans every stream for hype moments while analyzing every gameplay, video and audience behavior. The system then renders peak moments or “spikes,” calculates the relevant ranking and categorizes the type of each Spike.
Playstream has been working on advanced and innovative AI-based solutions for the past two years to help content creators grow by discovering new ways to monetize and gain audiences.
The product is a unique multidisciplinary AI solution, personalized and adaptable to the creator and their communities. Playstream’s solution scans every stream for hype moments while analyzing every gameplay, video and audience behavior. The system then renders peak moments or “spikes,” calculates the relevant ranking and categorizes the type of each Spike.
VentureBeat
Playstream raises $2M for AI-generated livestream content
Playstream has raised $2 million for its AI engine that generates the best gaming content from livestreams in seconds.
Why are people celebrating the Godfather-like economics of this OpenAI-Microsoft deal like it’s a good thing?
49% ownership with 33% of funds? with 75% cut of profits to EARN OUT the investment?
Was OpenAI strongarmed into this deal? wtf? We do not know the full story here.
The previous rumor was that they were raising 300m at 29b on conventional terms. I’m sure they could get 3b more from any number of late stage and xover funds.
OpenAI doesnt randomly need 10b and onerous deal structure for fun.
Either Semafor is dead wrong or all is not well.
49% ownership with 33% of funds? with 75% cut of profits to EARN OUT the investment?
Was OpenAI strongarmed into this deal? wtf? We do not know the full story here.
The previous rumor was that they were raising 300m at 29b on conventional terms. I’m sure they could get 3b more from any number of late stage and xover funds.
OpenAI doesnt randomly need 10b and onerous deal structure for fun.
Either Semafor is dead wrong or all is not well.
Semafor
Microsoft eyes $10 billion bet on ChatGPT
The tech giant has been in talks on deal to effectively own almost half of OpenAI, maker of the addictive, humanoid, AI-powered chatbot.
⚡️ New manuscript shows how zero-shot Generative AI can create de novo antibodies from scratch.
Hundreds of antibodies are created zero-shot and validated in the wet-lab for the first time ever.
Zero-shot generative AI framework. The model is programmed with a target antigen structure and a chosen antibody scaffold sequence.
Antibodies are then generated de novo.
Note: all proteins binding the target (or its homologs) were removed from the training set.
Example AI-generated cancer drug leads. A model generates antibodies with stronger affinities than a highly optimized therapeutic drug.
The binders come straight out of the model.
In some cases, >90% of the CDR3 region has changed.
Generative models have a tendency to memorize the training set. However, a model generates HCDR3s that are very different than those observed during training (on average, 5/13 positions are changed).
The sequences also differ from known antibodies in massive databases.
Paper
Data
Hundreds of antibodies are created zero-shot and validated in the wet-lab for the first time ever.
Zero-shot generative AI framework. The model is programmed with a target antigen structure and a chosen antibody scaffold sequence.
Antibodies are then generated de novo.
Note: all proteins binding the target (or its homologs) were removed from the training set.
Example AI-generated cancer drug leads. A model generates antibodies with stronger affinities than a highly optimized therapeutic drug.
The binders come straight out of the model.
In some cases, >90% of the CDR3 region has changed.
Generative models have a tendency to memorize the training set. However, a model generates HCDR3s that are very different than those observed during training (on average, 5/13 positions are changed).
The sequences also differ from known antibodies in massive databases.
Paper
Data
Forbes
This Company Is Using Generative AI To Design New Antibodies
GenerativeAI: You’ve seen it with images like DALL-E, you’ve seen it with text like ChatGPT. Now you can see it with protein design as well.
Timeline of OFAC crypto-related sanctions designations, 2018-2022.
Lessons_From_Crypto_Winter_1673437393.pdf
2.6 MB
the OECD - OCDE concluded in its latest report analyzing recent macro tradfi and crypto events, that there is an "urgent need for policy action" on crypto assets and that "a future instance of similar turmoil in a larger crypto-asset market could have implications for financial stability."