Apple just recently announced its first carbon neutral products – the new Apple Watch lineup. This achievement comes from innovations across Apple’s global supply chain over years to dramatically reduce emissions. It’s a major milestone toward Apple’s 2030 goal to make all products carbon neutral.
To become carbon neutral, Apple steeply cut watch emissions first via clean energy, recycled materials, and low-emission transportation. Any remaining emissions are addressed with high-quality carbon credits from nature-based projects like forests.
This shift demonstrates how companies can decarbonize operations and products through renewable electricity, material innovation, and carbon removal. If adopted widely, these strategies can significantly benefit the environment.
Apple’s progress was enabled by large investments in wind and solar energy. Their actions helped create over 15 gigawatts of new clean power. Scaling renewable energy is crucial for the transition away from fossil fuels.
The company also pioneered using recycled metals and fibers in devices. This reduces the need for carbon-intensive mining and materials manufacturing. Broad adoption would lessen impacts on natural resources.
Additionally, Apple funded carbon removal through forest restoration. This supports nature-based solutions to sequester CO2. The climate impact could grow exponentially if more firms financed conservation projects.
In summary, Apple’s carbon neutral product milestone highlights the environmental promise of renewable energy, the circular economy, and carbon removal. It demonstrates the potential for these strategies to transform manufacturing, conserve natural resources, and fight climate change.
The U.S. government is launching a monumental legal challenge against Google in a bid to curb the technology giant’s dominance in internet search. A federal antitrust trial begins Tuesday in Washington D.C. where the Justice Department and a coalition of state attorneys general will argue that Google improperly wields monopoly power.
At the heart of the case are allegations that Google unlawfully maintains its position in the search market through exclusionary distribution agreements and other anticompetitive practices. Google pays billions annually to companies like Apple and Samsung to preset Google as the default search engine on smartphones and other devices. This boxes out rivals, according to prosecutors.
The government contends that Google’s actions have suffocated competition in the critical gateway to the internet, enabling the company to extend its grasp with impunity. Google counters that its search supremacy is earned by offering a superior product that consumers freely choose, not due to illegal activity.
But smaller search upstarts like DuckDuckGo allege that Google abuses its might to hinder their ability to gain users. At stake in the trial is nothing less than how the power of dominant tech platforms is regulated and how competition – or lack of it – shapes the internet as we know it.
The verdict could lead to sweeping changes for Google if found guilty of violating antitrust law. Potential sanctions range from imposed restrictions on its business conduct to structural reorganization of the company. Fines could also be on the table.
Google’s practices echo the behavior that got Microsoft into hot water in the 1990s. That landmark case saw the government successfully prove Microsoft leveraged its Windows monopoly to quash competition. Google is accused of similar monopolistic plays via its search engine dominance.
The Google antitrust trial is slated to last around three months. Testimony from Google CEO Sundar Pichai and executives of tech firms like Apple is anticipated. The federal judge overseeing the case will determine if Google’s undisputed leadership in search equates to unlawful monopoly status.
The verdict stands to fundamentally shape Google’s role in internet search and potentially alter business practices of other dominant technology companies. It represents the most significant legal challenge to Silicon Valley power in the 21st century.
Online grocery delivery firm Instacart is gearing up to go public and has set the terms for its initial public offering (IPO). In a regulatory filing on Monday, Instacart outlined plans to raise around $616 million through the offering of 22 million shares priced between $26 and $28 each.
The IPO would give Instacart a fully diluted valuation of up to $9.3 billion. This is below earlier estimates of a $40 billion valuation, indicating moderating growth expectations. Nonetheless, the offering could still mark one of the largest public listings this year amid a freeze on IPOs over the past year due to market volatility.
Founded in 2012, San Francisco-based Instacart has established itself as a leading online grocery platform in the U.S. It partners with grocers and retailers to deliver items to customers’ doors in as little as an hour. Instacart competes in a crowded space against entrenched firms like Walmart and Amazon as well as delivery apps like DoorDash and GoPuff.
Take a moment to look at 1-800 Flowers.com, a leading e-commerce business platform that delivers gifts designed to help inspire customers to give more, connect more, and build more relationships.
Instacart plans to sell 14.1 million newly issued shares in the IPO, with the remainder offered by existing shareholders. Multiple prominent investors have committed to buying shares in the offering, including PepsiCo, which is investing $175 million, and Norges Bank Investment Management, Norway’s sovereign wealth fund.
Proceeds from the IPO will provide funding for Instacart to invest in areas like technology, fulfillment, and advertising as it aims to turn a profit. The company posted revenues of $1.8 billion in 2020 but has yet to become profitable.
The upcoming listing will test investor appetite for high-growth tech IPOs after a yearlong freeze. Instacart’s debut performance will depend on prevailing market sentiment closer to its trading date. But a successful IPO could boost Instacart’s brand and validate its status as a leading next-generation grocery platform.
Intercontinental Exchange (ICE), the financial markets data and infrastructure company, has finalized its $11.9 billion acquisition of Black Knight, a leading provider of mortgage software, data and analytics solutions.
The deal expands ICE’s growing footprint in mortgage technology services. Black Knight strengthens ICE’s capabilities spanning mortgage origination, servicing, and secondary market activities.
ICE, with a market valuation of $63 billion, has been actively acquiring assets to build out its mortgage tech segment. Previous deals include Ellie Mae, Simplifile and MERS. Black Knight, currently valued at around $10 billion, offers software and data services used by mortgage lenders, servicers, and real estate industry participants.
The combination aims to improve automation and digitization across the mortgage process through ICE’s financial resources and Black Knight’s housing domain expertise.
Black Knight shareholders could elect to receive the deal consideration in cash or ICE stock, subject to proration procedures. Preliminary results indicate strong demand for the stock option.
To secure regulatory clearances, ICE agreed to divest Black Knight’s Optimal Blue and Empower mortgage origination system businesses to Constellation Software Inc. “Our team is ready to apply our proven playbook to help improve the homeownership experience for millions of families,” said ICE CEO Jeffrey Sprecher.
The deal expands ICE’s information services and market infrastructure footprint into the massive U.S. housing market, while providing Black Knight greater scale and distribution capabilities.
Take a moment to learn about Information Services Group, a leading technology research and advisory firm that specializes in digital transformation services, including automation, cloud and data analytics, and market intelligence.
SoftBank Group’s Arm is gearing up for its highly-anticipated initial public offering (IPO), with ambitions to secure a valuation exceeding $52 billion. In an announcement made on Tuesday, the renowned chip designer unveiled plans to issue 95.5 million American depository shares, priced between $47 and $51 each, with a target of raising up to $4.87 billion at the upper end of this range.
While this valuation marks a decline from the $64 billion that SoftBank paid last month to acquire the remaining 25% stake in Arm from its $100 billion Vision Fund, it still surpasses the abandoned $40 billion sale of Arm to Nvidia Corp, which fell through last year due to opposition from antitrust regulators.
Arm, headquartered in Cambridge, England, holds a dominant position in the global technology landscape, powering over 99% of the world’s smartphones. Its innovative designs are also integral to a wide array of devices, spanning from tablets and laptops to servers and automobiles. Notably, Arm maintains a substantial presence in the United States.
Expected to be the largest IPO in the United States this year, Arm’s public offering carries significant weight as a litmus test for an IPO market grappling with challenges such as rising interest rates and geopolitical tensions stemming from the Ukraine conflict.
Despite these obstacles, investors are likely to welcome Arm’s IPO with open arms. The company boasts profitability and a remarkable history of technological innovation. Furthermore, Arm’s designs play a pivotal role in advancing emerging technologies like artificial intelligence and the metaverse.
For SoftBank, this IPO represents a major triumph. The Japanese conglomerate has been under pressure to enhance its investment returns, and while the sale of Arm would have been a monumental windfall, the IPO is a noteworthy achievement in its own right.
The success of Arm’s IPO hinges on several key factors:
1. IPO Market Conditions: The strength of the overall IPO market will play a vital role in determining Arm’s success.
2. Investor Appetite for Tech Stocks: As a technology company, Arm’s fate will be closely tied to investor sentiment towards tech stocks.
3. Valuation of Arm: The company’s valuation must be attractive to prospective investors.
4. Demand for Arm’s Shares: The level of demand for Arm’s shares will significantly impact the outcome.
If Arm’s IPO prevails, it could usher in a new era for the IPO market, potentially inspiring other startups to pursue public offerings. This success story would also bolster SoftBank’s financial standing and burnish its reputation as a savvy investor. Moreover, the technology industry would reap the rewards of heightened visibility and liquidity associated with Arm’s shares.
However, should Arm’s IPO falter, it could stymie the company’s growth prospects due to a lack of capital infusion. SoftBank would bear the financial brunt, and its reputation as an investor might suffer. Additionally, the technology sector would miss out on the potential benefits of Arm’s IPO.
In conclusion, Arm’s IPO is a watershed moment poised to leave an indelible mark on the company, SoftBank, and the technology sector at large. Its success will pivot on a complex interplay of factors, but if it prospers, it promises significant advantages for all stakeholders involved.
iCoreConnect, Inc. (Nasdaq: ICCT) recently underwent a business merger with FG Merger Corp. (Nasdaq: FGMC) and has since exhibited stability in the stock market. A notable event was a temporary halt in trading on Nasdaq due to a technical issue with the conversion of shares. However, trading resumed on August 30, 2023, after the issue was addressed. iCoreConnect is currently trending on various financial social media platforms and websites, reflecting heightened investor interest. Their stock price is up 206% since the start of the week as trading opened Friday.
iCoreConnect’s primary objective is to improve workflow productivity and practice profitability via its cloud-based software and technology solutions. Currently, the company has a portfolio of 16 SaaS enterprise solutions. Additionally, they’ve secured endorsements from over 100 state or regional healthcare associations in the U.S. Based on their recent statements, iCoreConnect has projected its revenue and annualized recurring revenue for 2023 and expressed interest in expanding into the ePrescription and insurance verification sectors.
To understand more about iCoreConnect’s activities, developments, and potential in the healthcare technology and enterprise solutions industry, a recent report from Noble Capital Markets Analyst Gergory Aurand provides a detailed analysis and overview.
Image: President Jimmy Carter and Chinese Vice Premier Deng Xiaoping meet outside of the Oval Office on Jan. 30, 1979
The US and China May Be Ending an Agreement on Science and Technology Cooperation − A Policy Expert Explains What This Means for Research
A decades-old science and technology cooperative agreement between the United States and China expires this week. On the surface, an expiring diplomatic agreement may not seem significant. But unless it’s renewed, the quiet end to a cooperative era may have consequences for scientific research and technological innovation.
The possible lapse comes after U.S. Rep. Mike Gallagher, R-Wis., led a congressional group warning the U.S. State Department in July 2023 to beware of cooperation with China. This group recommended to let the agreement expire without renewal, claiming China has gained a military advantage through its scientific and technological ties with the U.S.
The State Department has dragged its feet on renewing the agreement, only requesting an extension at the last moment to “amend and strengthen” the agreement.
The U.S. is an active international research collaborator, and since 2011 China has been its top scientific partner, displacing the United Kingdom, which had been the U.S.‘s most frequent collaborator for decades. China’s domestic research and development spending is closing in on parity with that of the United States. Its scholastic output is growing in both number and quality. According to recent studies, China’s science is becoming increasingly creative, breaking new ground.
This article was republished with permission from The Conversation, a news site dedicated to sharing ideas from academic experts. It represents the research-based findings and thoughts of,Caroline Wagner, Professor of Public Affairs, The Ohio State University.
As a policy analyst and public affairs professor, I research international collaboration in science and technology and its implications for public policy. Relations between countries are often enhanced by negotiating and signing agreements, and this agreement is no different. The U.S.’s science and technology agreement with China successfully built joint research projects and shared research centers between the two nations.
U.S. scientists can typically work with foreign counterparts without a political agreement. Most aren’t even aware of diplomatic agreements, which are signed long after researchers have worked together. But this is not the case with China, where the 1979 agreement became a prerequisite for and the initiator of cooperation.
In 1987 former President Jimmy Carter visited Yangshuo, his wife Rosalyn and he insisted that went around Yangshuo countryside by bicycle.
A 40-Year Diplomatic Investment
The U.S.-China science and technology agreement was part of a historic opening of relations between the two countries, following decades of antagonism and estrangement. U.S. President Richard Nixon set in motion the process of normalizing relations with China in the early 1970s. President Jimmy Carter continued to seek an improved relationship with China.
China had announced reforms, modernizations and a global opening after an intense period of isolation from the time of the Cultural Revolution from the late 1950s until the early 1970s. Among its “four modernizations” was science and technology, in addition to agriculture, defense and industry.
While China is historically known for inventing gunpowder, paper and the compass, China was not a scientific power in the 1970s. American and Chinese diplomats viewed science as a low-conflict activity, comparable to cultural exchange. They figured starting with a nonthreatening scientific agreement could pave the way for later discussions on more politically sensitive issues.
On July 28, 1979, Carter and Chinese Premier Deng Xiaoping signed an “umbrella agreement” that contained a general statement of intent to cooperate in science and technology, with specifics to be worked out later.
In the years that followed, China’s economy flourished, as did its scientific output. As China’s economy expanded, so did its investment in domestic research and development. This all boosted China’s ability to collaborate in science – aiding their own economy.
Early collaboration under the 1979 umbrella agreement was mostly symbolic and based upon information exchange, but substantive collaborations grew over time.
A major early achievement came when the two countries published research showing mothers could ingest folic acid to prevent birth defects like spina bifida in developing embryos. Other successful partnerships developed renewable energy, rapid diagnostic tests for the SARS virus and a solar-driven method for producing hydrogen fuel.
Joint projects then began to emerge independent of government agreements or aid. Researchers linked up around common interests – this is how nation-to-nation scientific collaboration thrives.
Many of these projects were initiated by Chinese Americans or Chinese nationals working in the United States who cooperated with researchers back home. In the earliest days of the COVID-19 pandemic, these strong ties led to rapid, increased Chinese-U.S. cooperation in response to the crisis.
Time of Conflict
Throughout the 2000s and 2010s, scientific collaboration between the two countries increased dramatically – joint research projects expanded, visiting students in science and engineering skyrocketed in number and collaborative publications received more recognition.
As China’s economy and technological success grew, however, U.S. government agencies and Congress began to scrutinize the agreement and its output. Chinese know-how began to build military strength and, with China’s military and political influence growing, they worried about intellectual property theft, trade secret violations and national security vulnerabilities coming from connections with the U.S.
Recent U.S. legislation, such as the CHIPS and Science Act, is a direct response to China’s stunning expansion. Through the CHIPS and Science Act, the U.S. will boost its semiconductor industry, seen as the platform for building future industries, while seeking to limit China’s access to advances in AI and electronics.
A Victim of Success?
Some politicians believe this bilateral science and technology agreement, negotiated in the 1970s as the least contentious form of cooperation – and one renewed many times – may now threaten the United States’ dominance in science and technology. As political and military tensions grow, both countries are wary of renewal of the agreement, even as China has signed similar agreements with over 100 nations.
The United States is stuck in a world that no longer exists – one where it dominates science and technology. China now leads the world in research publications recognized as high quality work, and it produces many more engineers than the U.S. By all measures, China’s research spending is soaring.
Even if the recent extension results in a renegotiated agreement, the U.S. has signaled to China a reluctance to cooperate. Since 2018, joint publications have dropped in number. Chinese researchers are less willing to come to the U.S. Meanwhile, Chinese researchers who are in the U.S. are increasingly likely to return home taking valuable knowledge with them.
The U.S. risks being cut off from top know-how as China forges ahead. Perhaps looking at science as a globally shared resource could help both parties craft a truly “win-win” agreement.
What History Says About MegaCap Companies with NVDA’s Multiples
What’s riskier, a stock like Nvidia that has been moving up since the start of the year and has now risen more than 200%, or the stock you pick by throwing a dart at the Wall Street Journal? Nvidia (NVDA) will report earnings on Wednesday, the report has a significant risk of disappointing, even if it exceeds forecasts.
Background
The last time Nvidia reported quarterly results, the chip maker forecasted record revenue that was far above anything it had accomplished before. Naturally, investors sent the stock up and then up some more to the level we see today. On Wednesday, August 23rd, the company will share the actual results of its second-quarter earnings. If it doesn’t deliver or exceed expectations, there may be a lot of disappointed investors. Plus, a repeat projection of future earnings growth would seem necessary to maintain its trajectory. Nvidia’s current valuation is extremely high, it has been the poster child of the AI investment boom, it would seem there are more scenarios where it will disappoint the frenzy in the market for the stock then there are positives for a company with such high valuations.
The phrase “the bigger they are, the harder they fall” was used in the context of the stock market as early as the 19th century. In 1873, the New York Stock Exchange crashed, and many investors lost their life savings. The crash was blamed on the overvaluation of stocks. Since then, there have been a number of times the market and individual stocks have come to terms with the idea that the price has gotten ahead of itself, causing speculators to flee, causing a sharp decline.
Nvidia was big and has gotten bigger. As it relates to its report on second-quarter earnings on Wednesday, analysts have been expressing positive expectations. In a “buy the rumor sell the news” way of thinking, this alone may cause the stock price to deflate after the report.
An article published by Barron’s on August 21st points out, using data from WisdomTree, that Nvidia has a price-to-projected sales ratio, of 25, this jumps to a whopping 40 when using 12-month trailing sales.
The company now holds the distinction of having the highest price-to-sales ratio in the S&P 500. It is only the 100th company to hold that title since the 1960s. Tech companies had been in this position 27 times, and utilities a mere once. What has happened to these mega-cap hot stocks in the years to follow?
Image: Research conducted for stocks for the long run 6th edition, Jeremy Siegel w/ Jeremy Schwartz (2022).
Over the next year, the average price-to-sales monsters saw their price rise 12% versus the average for the market of 11%. Then, over the next three years, the average stock dropped 4% annually, compared to the market’s 9% rise. The relative fall off continued over the next five years, as the average for these stocks had fallen 2%, versus the market’s 10% gain, according to a book by Wharton economic professor Jeremy Siegel.
Image: Research conducted for stocks for the long run 6th edition, Jeremy Siegel w/ Jeremy Schwartz (2022).
As illustrated in the table above, the underperformance is even greater for the tech companies that hold the biggest price-to-sales ratios. And the median performance for stocks at these price-to-sales ratios is even worse.
Take Away
Based on its guidance, there is little doubt that Nvidia is on track to post results that are beyond enviable. At least a dozen equity analysts have recently raised price targets on NVDA. But can the news be good enough in light of its current pricing and the history of tech stocks that have come before? NVDA would have been a great stock to have bought months ago and held; the current probabilities, based on history, now suggest the run-up is over, and the potential for a decline within a year has increased dramatically.
Bitcoin and Ethereum had a bad day. After gaining a lot of upward momentum from late June after Blackrock, Fidelity, and Invesco filed to create bitcoin-related exchange traded funds (ETFs), the volatile assets have shown cryptocurrency investors that the bumpy ride is not yet over. What’s causing it this time? Fortunately, it is not fraud or wrongdoing creating the turbulence. Instead, three factors external to the business of trading, mining, or exchanging digital assets are at work.
Background
On Thursday, August 17, and accelerating on August 18, the largest cryptocurrencies dropped precipitously. Bitcoin even broke down and fell below the psychologically important $26,000 US dollar price level before bouncing. While some are pointing to CME options expiration on the third Friday of each month, most are pointing to a Wall Street Journal article, and blaming Elon Musk, as the reason the asset class was nudged off a small cliff. There are other less highlighted, but important, catalysts that added to the flash-crash; these, along with the WSJ story, will be explained below.
Smells like Musk
What could SpaceX, the company owned and run by Elon Musk, possibly have to do with a crypto selloff? On Thursday, the crypto market had a downward spike around 5 PM ET. It was just after the Wall Street Journal revealed a change in the accounting valuation of SpaceX’s crypto assets. Reportedly, SpaceX marked down the value of its bitcoin assets by a substantial $373 million over the past two years. Additionally, the company has executed on crypto asset divestitures as well. When the reduction took place is uncertain, but cryptocurrency holdings have been reduced both in terms of the amount of coins and the value each coin is held for on the books.
Elon Musk’s reputation is that of a forward thinker, and one that embraces, if not leads, technology. He has significant influence over cryptocurrency valuations, often instigating pronounced market fluctuations brought about by Musk’s influential posts on his social media company, X. The reduction coincides with a similar crypto reduction on the books of publicly held, Musk-led, Tesla (TSLA). The electric car manufacturer had previously disclosed in its annual earnings report that it had liquidated 75% of its bitcoin reserves.
While it should not be surprising that two companies stepped away from speculation on something unrelated to their business or lowered support for the still young blockchain technology, it gave a reason for a reaction to this and other festering dynamics.
Wary of Gary
The Chairman of the Securities and Exchange Commission (SEC), Gary Gensler, is viewed as a “Whack-a Mole” to crypto stakeholders that prefer more autonomy than regulation. Every time the SEC gets knocked down as a potential regulator, it resurfaces, and crypto businesses have to deal with the agency again.
Last month, Judge Analisa Torres made a pivotal decision in a case involving payment company Ripple Labs and the Commission. Her verdict declared that a substantial portion of sales of the token XRP did not fall under the category of securities transactions. The SEC claimed it was a security. This judgement was hailed as a triumph for the crypto sector and catalyzed an impressive 20% uptick in the exchange Coinbase’s stock in a single day.
On the same Thursday as the WSJ article, the SEC showed its face again with a strong response to the earlier ruling. Judge Torres allowed the SEC’s request for an “interlocutory” appeal on her ruling. This process will involve the SEC presenting its motion, followed by Ripple’s counterarguments. This is slated to continue until mid-September. Afterward, the Judge will determine whether the agency can effectively challenge her token classification ruling in an appellate court.
The still young asset class, its exchange methods, valuation, and usage techniques, once they are more clearly defined, will serve to add stability and reduce risk and shocks in crypto and the surrounding businesses. The longer the legal system and regulatory entities take, including Congress, the longer it will take for cryptocurrencies to find the more settled mainstream place in the markets they desire.
Rate Spate
The eighteen-month-long spate of rate hikes in the U.S. and across the globe is providing an alternative investment choice instead of what are viewed as riskier assets. Coincidentally, again on Thursday, August 17, the ten-year US Treasury Note hit a yield higher than the markets have experienced in 12 years. At 4.31%, investors can lock in a known annual return for ten years that exceeds the current and projected inflation rate.
Take Away
The volatility in the crypto asset class has been dramatic – not for the weak-stomached investor. On the same day in August, three unrelated events together helped cause the asset class to spike down. These include an article in a top business news publication indicating that one of the world’s most recognized cryptocurrency advocates has reduced bitcoin’s exposure to his companies. The SEC being granted a rematch in a landmark case that it had recently lost, where the earlier outcome gave no provision for the SEC to treat cryptocurrencies like a security. And rounding out the triad of events on crypto’s throttleback Thursday, yields are up across the curve to levels not seen in a dozen years. Investor’s seeking a place to reduce risk can now provide themselves with interest payments in excess of inflation.
But despite the ups and downs, bitcoin is up 56.7% year-to-date, 11.1% over the past 12 months, 110.5% over three years, 300% over five years, and astronomical amounts over longer periods. Related companies like bitcoin miners, crypto exchanges, and blockchain companies have also experienced growth similar to that found in few other industries over the past decade.
Artificial neural networks, ubiquitous machine-learning models that can be trained to complete many tasks, are so called because their architecture is inspired by the way biological neurons process information in the human brain.
About six years ago, scientists discovered a new type of more powerful neural network model known as a transformer. These models can achieve unprecedented performance, such as by generating text from prompts with near-human-like accuracy. A transformer underlies AI systems such as ChatGPT and Bard, for example. While incredibly effective, transformers are also mysterious: Unlike with other brain-inspired neural network models, it hasn’t been clear how to build them using biological components.
Now, researchers from MIT, the MIT-IBM Watson AI Lab, and Harvard Medical School have produced a hypothesis that may explain how a transformer could be built using biological elements in the brain. They suggest that a biological network composed of neurons and other brain cells called astrocytes could perform the same core computation as a transformer.
Recent research has shown that astrocytes, non-neuronal cells that are abundant in the brain, communicate with neurons and play a role in some physiological processes, like regulating blood flow. But scientists still lack a clear understanding of what these cells do computationally.
With the new study, published this week in open-access format in the Proceedings of the National Academy of Sciences, the researchers explored the role astrocytes play in the brain from a computational perspective, and crafted a mathematical model that shows how they could be used, along with neurons, to build a biologically plausible transformer.
Their hypothesis provides insights that could spark future neuroscience research into how the human brain works. At the same time, it could help machine-learning researchers explain why transformers are so successful across a diverse set of complex tasks.
“The brain is far superior to even the best artificial neural networks that we have developed, but we don’t really know exactly how the brain works. There is scientific value in thinking about connections between biological hardware and large-scale artificial intelligence networks. This is neuroscience for AI and AI for neuroscience,” says Dmitry Krotov, a research staff member at the MIT-IBM Watson AI Lab and senior author of the research paper.
Joining Krotov on the paper are lead author Leo Kozachkov, a postdoc in the MIT Department of Brain and Cognitive Sciences; and Ksenia V. Kastanenka, an assistant professor of neurobiology at Harvard Medical School and an assistant investigator at the Massachusetts General Research Institute.
A Biological Impossibility Becomes Plausible
Transformers operate differently than other neural network models. For instance, a recurrent neural network trained for natural language processing would compare each word in a sentence to an internal state determined by the previous words. A transformer, on the other hand, compares all the words in the sentence at once to generate a prediction, a process called self-attention.
For self-attention to work, the transformer must keep all the words ready in some form of memory, Krotov explains, but this didn’t seem biologically possible due to the way neurons communicate.
However, a few years ago scientists studying a slightly different type of machine-learning model (known as a Dense Associated Memory) realized that this self-attention mechanism could occur in the brain, but only if there were communication between at least three neurons.
“The number three really popped out to me because it is known in neuroscience that these cells called astrocytes, which are not neurons, form three-way connections with neurons, what are called tripartite synapses,” Kozachkov says.
When two neurons communicate, a presynaptic neuron sends chemicals called neurotransmitters across the synapse that connects it to a postsynaptic neuron. Sometimes, an astrocyte is also connected — it wraps a long, thin tentacle around the synapse, creating a tripartite (three-part) synapse. One astrocyte may form millions of tripartite synapses.
The astrocyte collects some neurotransmitters that flow through the synaptic junction. At some point, the astrocyte can signal back to the neurons. Because astrocytes operate on a much longer time scale than neurons — they create signals by slowly elevating their calcium response and then decreasing it — these cells can hold and integrate information communicated to them from neurons. In this way, astrocytes can form a type of memory buffer, Krotov says.
“If you think about it from that perspective, then astrocytes are extremely natural for precisely the computation we need to perform the attention operation inside transformers,” he adds.
Building a Neuron-Astrocyte Network
With this insight, the researchers formed their hypothesis that astrocytes could play a role in how transformers compute. Then they set out to build a mathematical model of a neuron-astrocyte network that would operate like a transformer.
They took the core mathematics that comprise a transformer and developed simple biophysical models of what astrocytes and neurons do when they communicate in the brain, based on a deep dive into the literature and guidance from neuroscientist collaborators.
Then they combined the models in certain ways until they arrived at an equation of a neuron-astrocyte network that describes a transformer’s self-attention.
“Sometimes, we found that certain things we wanted to be true couldn’t be plausibly implemented. So, we had to think of workarounds. There are some things in the paper that are very careful approximations of the transformer architecture to be able to match it in a biologically plausible way,” Kozachkov says.
Through their analysis, the researchers showed that their biophysical neuron-astrocyte network theoretically matches a transformer. In addition, they conducted numerical simulations by feeding images and paragraphs of text to transformer models and comparing the responses to those of their simulated neuron-astrocyte network. Both responded to the prompts in similar ways, confirming their theoretical model.
“Having remained electrically silent for over a century of brain recordings, astrocytes are one of the most abundant, yet less explored, cells in the brain. The potential of unleashing the computational power of the other half of our brain is enormous,” says Konstantinos Michmizos, associate professor of computer science at Rutgers University, who was not involved with this work. “This study opens up a fascinating iterative loop, from understanding how intelligent behavior may truly emerge in the brain, to translating disruptive hypotheses into new tools that exhibit human-like intelligence.”
The next step for the researchers is to make the leap from theory to practice. They hope to compare the model’s predictions to those that have been observed in biological experiments, and use this knowledge to refine, or possibly disprove, their hypothesis.
In addition, one implication of their study is that astrocytes may be involved in long-term memory, since the network needs to store information to be able act on it in the future. Additional research could investigate this idea further, Krotov says.
“For a lot of reasons, astrocytes are extremely important for cognition and behavior, and they operate in fundamentally different ways from neurons. My biggest hope for this paper is that it catalyzes a bunch of research in computational neuroscience toward glial cells, and in particular, astrocytes,” adds Kozachkov.
Blackboxstocks, Inc. is a financial technology and social media hybrid platform offering real-time proprietary analytics and news for stock and options traders of all levels. Our web-based software employs “predictive technology” enhanced by artificial intelligence to find volatility and unusual market activity that may result in the rapid change in the price of a stock or option. Blackbox continuously scans the NASDAQ, New York Stock Exchange, CBOE, and all other options markets, analyzing over 10,000 stocks and up to 1,500,000 options contracts multiple times per second. We provide our users with a fully interactive social media platform that is integrated into our dashboard, enabling our users to exchange information and ideas quickly and efficiently through a common network. We recently introduced a live audio/video feature that allows our members to broadcast on their own channels to share trade strategies and market insight within the Blackbox community. Blackbox is a SaaS company with a growing base of users that spans 42 countries; current subscription fees are $99.97 per month or $959.00 annually. For more information, go to: www.blackboxstocks.com .
Joe Gomes, Managing Director, Equity Research Analyst, Generalist , Noble Capital Markets, Inc.
Joshua Zoepfel, Research Associate, Noble Capital Markets, Inc.
Refer to the full report for the price target, fundamental analysis, and rating.
2Q23. Revenue was $737,398, down from $1.4 million a year ago. Average monthly revenue per subscriber declined to $64.27 from $75.21 due to a higher level of promotional members during 2Q23. Blackboxstocks reported a net loss of $1.4 million for the quarter, or a loss of $0.45 per share, compared to a net loss of $1.3 million, or a loss of $0.40 per share in 2Q22. Adjusted EBITDA was a loss of $1.0 million, similar to the adjusted EBITDA loss in the year ago quarter.
Subscriber Counts. The average member count for the second quarter of 2023 was 3,937 compared to 5,482 for the second quarter of 2022 and 3,555 for the first quarter of 2023. The sequential stabilization of the member count reflects lower churn, while changes to marketing are expected to have a positive impact on the growth trajectory going forward.
Equity Research is available at no cost to Registered users of Channelchek. Not a Member? Click ‘Join’ to join the Channelchek Community. There is no cost to register, and we never collect credit card information.
This Company Sponsored Research is provided by Noble Capital Markets, Inc., a FINRA and S.E.C. registered broker-dealer (B/D).
*Analyst certification and important disclosures included in the full report. NOTE: investment decisions should not be based upon the content of this research summary. Proper due diligence is required before making any investment decision.
One Stop Systems, Inc. (OSS) designs and manufactures innovative AI Transportable edge computing modules and systems, including ruggedized servers, compute accelerators, expansion systems, flash storage arrays, and Ion Accelerator™ SAN, NAS, and data recording software for AI workflows. These products are used for AI data set capture, training, and large-scale inference in the defense, oil and gas, mining, autonomous vehicles, and rugged entertainment applications. OSS utilizes the power of PCI Express, the latest GPU accelerators and NVMe storage to build award-winning systems, including many industry firsts, for industrial OEMs and government customers. The company enables AI on the Fly® by bringing AI datacenter performance to ‘the edge,’ especially on mobile platforms, and by addressing the entire AI workflow, from high-speed data acquisition to deep learning, training, and inference. OSS products are available directly or through global distributors. For more information, go to www.onestopsystems.com.
Joe Gomes, Managing Director, Equity Research Analyst, Generalist , Noble Capital Markets, Inc.
Joshua Zoepfel, Research Associate, Noble Capital Markets, Inc.
Refer to the full report for the price target, fundamental analysis, and rating.
2Q23 Results. Revenue of $17.2 million, slightly below guidance and down 6% y-o-y, reflecting the continued runoff of the Disguise business. Higher operating expenses, including some one-time items, drove a net loss of $2.4 million, or a loss of $0.12/sh in the quarter, compared to net income of $322,822, or EPS of $0.02/sh per share last year. We had forecast a net loss of $415,700, or a loss of $0.02/sh. Adjusted EPS was $0.01 compared to $0.04 last year.
Delays and Departures. The decline in Disguise (substantially complete), delays in defense and commercial program orders, and the exit of an autonomous trucking client, a top 10 client in 2022, have combined to stall the momentum of OSS’ business. Revenue guidance for 3Q23 is $13.5 million, a level last seen in the first quarter of 2021.
Equity Research is available at no cost to Registered users of Channelchek. Not a Member? Click ‘Join’ to join the Channelchek Community. There is no cost to register, and we never collect credit card information.
This Company Sponsored Research is provided by Noble Capital Markets, Inc., a FINRA and S.E.C. registered broker-dealer (B/D).
*Analyst certification and important disclosures included in the full report. NOTE: investment decisions should not be based upon the content of this research summary. Proper due diligence is required before making any investment decision.
ISG (Information Services Group) (Nasdaq: III) is a leading global technology research and advisory firm. A trusted business partner to more than 700 clients, including more than 75 of the world’s top 100 enterprises, ISG is committed to helping corporations, public sector organizations, and service and technology providers achieve operational excellence and faster growth. The firm specializes in digital transformation services, including automation, cloud and data analytics; sourcing advisory; managed governance and risk services; network carrier services; strategy and operations design; change management; market intelligence and technology research and analysis. Founded in 2006, and based in Stamford, Conn., ISG employs more than 1,300 digital-ready professionals operating in more than 20 countries—a global team known for its innovative thinking, market influence, deep industry and technology expertise, and world-class research and analytical capabilities based on the industry’s most comprehensive marketplace data. For additional information, visit www.ISG-One.com
Joe Gomes, Managing Director, Equity Research Analyst, Generalist , Noble Capital Markets, Inc.
Joshua Zoepfel, Research Associate, Noble Capital Markets, Inc.
Refer to the full report for the price target, fundamental analysis, and rating.
Still Seeing Demand. ISG is continuing to experience increasing demand for its services, evidenced by the record second quarter revenue of $74.6 million. As noted in our previous report, recurring revenues now represents more than 40% of Company revenue and we anticipate the number to climb due to further demand from verticals such as consumer services, life sciences, and the public sector for the Company’s cost optimization services.
Balance Sheet. ISG reported a cash balance of $19.6 million as of June 30, 2023 compared to $23.7 million at March 31, 2023. The Company’s debt outstanding was at $79.2 million, unchanged from the first quarter, with the Company’s debt-to-EBITDA ratio remaining at 1.8x.
Equity Research is available at no cost to Registered users of Channelchek. Not a Member? Click ‘Join’ to join the Channelchek Community. There is no cost to register, and we never collect credit card information.
This Company Sponsored Research is provided by Noble Capital Markets, Inc., a FINRA and S.E.C. registered broker-dealer (B/D).
*Analyst certification and important disclosures included in the full report. NOTE: investment decisions should not be based upon the content of this research summary. Proper due diligence is required before making any investment decision.
