July 2023 Newsletter
1. AI - COPYRIGHT INFRINGEMENT
Sarah Silverman Sues OpenAI, Meta for Copyright Infringement?
July 10, 2023: In a recent development, comedian Sarah Silverman has filed a lawsuit against OpenAI and Meta, formerly known as Facebook, alleging copyright infringement. The lawsuit, filed in the U.S. District Court for the Northern District of California, accuses the two tech giants of using her jokes without permission.
Silverman's legal team argues that OpenAI's language model, GPT-3, and Meta's AI system have been programmed to generate jokes strikingly similar to those in her comedy routines. The comedian claims these AI systems have been using her copyrighted material, infringing her rights.
The lawsuit is a significant development in the ongoing debate about the intersection of AI and copyright law. It raises important questions about whether AI systems can infringe on copyright by generating content similar to existing copyrighted works.
2. US AGENCIES - PERSONAL DATA
US agencies buy personal information on the open market
June 29, 2023: Here's something that might surprise you . Numerous U.S. government agencies, including the FBI, Department of Defense, and National Security Agency, are buying up personal information. This revelation, from a partially declassified report, shines a light on the scale and potential of the consumer data market.
The report highlights the collection of various types of information such as location, gender and sexual orientation, religious and political views and affiliations, weight and blood pressure, speech patterns, emotional states, behavioral information, shopping patterns, and family and friends. It emphasizes the significant dangers of acquiring this data and recommends that the intelligence community establish internal protocols to tackle these issues.
There is no legal prohibition on the government collecting information already disclosed to the public or otherwise publicly available. But the nonpublic information listed in the declassified report includes data that U.S. law typically protects. The nonpublic information’s mix of private, sensitive, confidential, or otherwise lawfully protected data makes the collection a legal gray area.
3. GENERATIVE AI
Databricks Strikes $1.3 Billion Deal for Generative AI Startup MosaicML?
June 26, 2023: Databricks’ decision to acquire MosaicML addresses a challenging concern: developing LLMs for artificial intelligence(AI) at a lower cost point and with fewer compute resources. Mosaic helps organizations develop their LLMs based on corporate data, not publicly accessing content. The move by Databricks also helps lower the cost for organizations to create and utilize their own LLMs versus licensing existing LLMs.?
Databricks’ strategy incorporates extending the client's ability to create 15 times lower-cost LLMs within the Lakehouse environment. With the announcement of LakeHouseIQ, this platform will bring together Databricks Lakehouse, MosaicML, and the Unity catalog as one offering. Like other LLM creators, Databricks will rent this platform to their clients.?
The cost for licensing LLMs today could range in the millions of dollars. MosaicML helps drive down the cost of inference and training, and the process will also require fewer hardware platforms by optimizing GPU technology.?
This hardware offload efficiency gives Databricks several advantages compared to open-source and closed platforms. With access to a lower LLM capability based on the organization’s data, the learning process becomes more efficient with fewer data management requirements at the process's inception; Databricks extends the process of fine-tuning in their offering.?
Organizations tapping their data into their LLM process could cut costs by 50% or more. More importantly, organizations moving ahead with AI and ML across several areas within their enterprise will have more confidence in security and privacy because the LLMs become based on their data.?
June 15, 2023: IT leaders are uncertain about their organization’s ability to support AI’s growing use and demands.
The 2023 Global Tech Trends Survey by Equinix reported that merely 58% of global IT leaders (68% in the Americas) feel “very comfortable” with their “organization’s infrastructure [and] team’s ability to accommodate AI’s growing use.” IT leaders reported worrying most about increasing operating costs, insufficient internal knowledge, and slow speed of implementing new technologies like AI.
Still, most IT leaders in the Americas claimed that they currently use or will use AI for IT operations (91%) and cybersecurity (85%). Other top functions for AI included customer experience, research & development, and marketing. The survey respondents included 2,900 global IT decision-makers.
IT leaders are asked to support advanced technologies like AI despite competing priorities. That said, the AI vendor ecosystem is maturing, which can lower barriers to adoption for some enterprises.
June 14, 2023: Google has introduced a new feature that utilizes AI to provide shoppers a virtual try-on experience, allowing them to visualize how clothing items would look on various models. The system enables users to select the "Try On" badge on Search and choose a model that closely matches their body type—addressing the common challenge faced by online shoppers who cannot physically try on garments and must rely on photos of models, which may not accurately represent their own body shape, skin tone, or hair color.
The recent development by Google enables online shoppers to make more informed purchasing decisions. The virtual try-on feature employs TryOnDiffusion, a generative AI system trained by Google researchers to analyze clothing images and predict how they appear on various models, spanning from XXS to 4XL sizes and showcasing multiple poses.
While Google is not the first company to offer virtual try-on capabilities, its approach distinguishes it from competitors like Walmart. Walmart's virtual try-on feature only allows users to view clothes on models with similar height and size or photos of themselves. However, Google's technology surpasses this by offering a broader range of poses and a more comprehensive selection of models.?
The virtual try-on feature is made possible through a diffusion-based AI model developed internally by Google. This model gradually refines noisy images by subtracting noise and moving closer to a target image. By training the model using pairs of images featuring individuals wearing garments in different poses, Google ensures that the virtual try-on experience produces realistic results, considering factors such as fabric stretch and wrinkles.
Overall, Google's virtual try-on feature, powered by AI, empowers online shoppers to make more informed decisions by visualizing clothing items on models that resemble their body types. As part of its broader strategy, Google aims to enhance the shopping experience and capture a larger share of the e-commerce market while addressing concerns regarding diversity in the fashion industry.
Google's new feature brings advantages to shoppers, yet it also sparks concerns about its potential consequences for models in the fashion industry. Critics argue that Generative AI may intensify inequalities, such as limited model diversity. To tackle these concerns, Google strives to incorporate real models of varying sizes, ethnicities, skin tones, body shapes, and hair types in its virtual try-on feature.
The introduction of this AI-powered feature aligns with Google's broader strategy of becoming a comprehensive shopping platform and connecting merchants with consumers. Google aims to leverage its open ecosystem and the power of AI to enhance the shopping experience. This move comes as a response to the growing competition from e-commerce giants like Amazon and emerging platforms such as TikTok, attracting younger audiences.
In addition to the virtual try-on feature, Google is introducing various updates to Google Shopping. These updates include AI-powered filtering options for clothing searches, enabling users to refine their searches based on criteria like color, style, and pattern. Furthermore, Google is expanding its use of generative AI to enhance other areas, such as travel destination research and mapping routes.
ADDITIONAL SOURCES
6. ORACLE - GENERATIVE AI
Oracle will collaborate with AI startup and Open AI rival to develop generative AI services?
June 13, 2023: Oracle Plans to Provide Generative AI Services with Cohere
Leading tech companies are all heavily investing in generative AI technologies. Google has invested $250 million into Anthropic, Microsoft has invested $10 billion in OpenAI, and most recently, Oracle has joined the race. Oracle announced its partnership with Cohere in June, only a week after participating in Cohere’s $270 million funding round.
Cohere specializes in foundation models–a system that trains on large quantities of unstructured data that businesses can apply to many functions. By partnering with Oracle, they will have access to its vast database market and customer base. In exchange, Oracle will be able to bring AI products to market in a fraction of the time while diversifying the LLM space with a different provider.
Yet, this partnership also exposes Oracle to the same challenges that other vendors in this space experience. Governance, security, and privacy issues are still rife within AI. Oracle has countered this by suggesting that their OCI’s generative AI capabilities will give customers more control over their data.
Results from the partnership are already on the horizon, with the Oracle Fusion Cloud Human Capital Management (HCM) HR software offering AI integrations. When an HR agent needs to create a job listing or list performance goals, Oracle Fusion Cloud HCM will contain a button that automatically generates these fields. These integrations will launch by the end of the year.
Oracle attributes its rapid progress to the superiority of OCI’s generative workload management in the cloud. It combines apps and infrastructure to deliver faster, cheaper, and more integrated processes. With its new partnership, Oracle is leaning into the surge of AI technology the entire sector is experiencing.
Additional Sources
7. ASSET MANAGEMENT
Around A Third of Enterprise Software Spend Is Wasted
June 13, 2023: Organizations waste or underutilize over 32% of their IT budgets , according to Flexera’s 2023 State of ITAM (IT Asset Management) report. Their report took place in March of 2023 and surveyed 500 technical professionals and executive leaders across the globe. The report states that organizations waste 36% of desktop software spend, 33% of data center software spend, 32% of SaaS spend, and 32% of IaaS/PaaS spend.
Organizations should incorporate a SAM (Software Asset Management) program to increase IT savings and decrease wasted expenditure. In the past year since creating a SAM program, 55% of respondents have saved over $1 million. A further 16% have saved over $10 million, with 7% of that group saving over $50 million due to their SAM program.?
SAM platforms optimize software use to save money, increase the maturity of the software supply chain, deal with SaaS containers, and reduce the complexity of vendor use rights. Savings from SAM programs tend to increase as programs mature. The main benefit of a beginner program is better negotiation of vendor contracts. In contrast, advanced programs mainly encounter methods of reusing non-cloud licenses to mitigate the need to buy new ones.
Another aspect of SAM programs that makes them efficient is their high level of interaction with other department teams. Across FinOps, CIO/CTO, Security, Infrastructure, and IT Service Management, SAM programs have a significant level of interaction. For modern organizations that find themselves in a nexus of FinOps, security, ESG, cost savings, and other enterprise initiatives, this allows SAM programs to help save across the board.
Over the past year, there has been an increase in visibility into an organization's IT estate. About 39% of organizations now have this visibility, which is a slight improvement from the previous year's 35%. However, that still suggests that 61% of respondents don’t have complete visibility into IT assets and cannot effectively monitor business outcomes.?
SAM programs and other ITAM tools can increase savings and boost visibility for most organizations. Yet, there is still a long way to go to reduce corporate inefficiencies.?
Additional Sources
8. BLOCKCHAIN - AI
AI x Blockchain: The Next Level?
June 13, 2023: Blockchain and artificial intelligence (AI) are two emerging technologies that have had a major impact over the last decade. At the Coinbase Machine Learning (ML) and Blockchain Summit, four leaders from academia and industry came together to explore the crossroads of these technologies and the future promise they hold.
领英推荐
The panel outlined how ML models can analyze transaction data to expose potential misconduct on blockchain networks and detect emerging security threats. AI could dynamically adjust transaction fees to optimize system resources based on current trading volumes.
Source .
By integrating on-chain data from blockchain and off-chain internet data, ML could expose blockchain systems to Web 2 data, creating a new wave of potential. This integration would allow blockchains to expand beyond their self-contained systems, creating the opportunity to build more dynamic ecosystems.?
However, perhaps the greatest benefit of combining these technologies would be to streamline, optimize, and increase the security of AI tools.
Blockchain is a decentralized, transparent, and auditable system. Due to these qualities, it could at least partially solve rising concerns about AI systems' privacy, bias, and security. Blockchain could actively counter the misinformation within AI datasets, using cryptographic signatures and timestamps to authenticate the information that AI utilizes.
Using data from the blockchain could accelerate AI development and streamline the production of LLMs (Large Language Models). Additionally, companies could create borderless, internet-native payment systems with blockchain, providing financial incentives to users that submit their data to LLMs.?
All the experts agree that using blockchain and AI technologically symbiotically has the potential to advance both fields. However, they also suggest that businesses must put users’ needs for safety and privacy first as innovation continues.?
Additional Sources
9. SALESFORCE - GENERATIVE AI CLOUD
Salesforce announces AI Cloud to empower enterprises with trusted generative AI
June 12, 2023: Salesforce has introduced AI Cloud, a solution designed to enhance productivity across all Salesforce applications, catering specifically to enterprise needs. This innovative open platform seamlessly combines multiple Salesforce technologies, including Einstein, Data Cloud, Tableau, Flow, and MuleSoft, enabling real-time generative AI capabilities to optimize business operations.
AI Cloud's Einstein Trust Layer enhances trust in enterprise generative AI by safeguarding sensitive data, preventing the integration of proprietary information into public models, and addressing concerns about data privacy, security, residency, and compliance. Customers can also select customized large language models (LLMs) from providers like OpenAI, Anthropic, Cohere, and Salesforce's exclusive proprietary and domain-specific models.
Salesforce's LLMs, including CodeGen, CodeT5+, and CodeTF, offer advanced capabilities like code generation and business process automation assistance.
"We created AI Cloud to help customers leverage AI-created content across sales, customer service, marketing, commerce and IT interactions to better connect with their audiences in new, more personalized ways," said Adam Caplan, SVP of emerging technology at Salesforce.
"Our new offering will empower enterprises to safely turn on the power of generative AI across all of their applications, connect all of their enterprise data, gain new insights about their customers, and automate workflows."
The company has also developed prompt templates and builders to optimize workflows and improve the quality and relevance of AI-generated content. These tools use harmonized data to ground the generated outputs in each company's unique context and reduce or eliminate AI hallucinations.
The new offering will enable sales teams to generate personalized emails customized to meet customers' needs. Additionally, marketers can use the tool to create audience segments, using natural language prompts and AI-driven recommendations to enhance targeting.??????????????????????????
"Our guidelines for responsible generative AI aim to help Salesforce and all users of generative AI mitigate these issues by following responsible best practices when it comes to development," said Caplan.?
10. BLOCKCHAIN - INDIA
JPMorgan turns to Blockchain for dollar trades in India hub?
June 5, 2023: JPMorgan Chase & Co. has partnered with six Indian banks in India to settle interbank dollar transactions in a growing national commerce hub. Alongside HDFC Bank, ICICI Bank, Axis Bank, Yes Bank, IndusInd Bank, and JPMorgan's unit, they will establish a blockchain-based platform that settles interbank dollar transactions.?
The existing system processes settlements in hours and does not function on weekends or public holidays. By leveraging blockchain technology, the system will now be able to complete transactions instantly on a 24/7 basis. The platform will launch on Onyx, JPMorgan's blockchain-based system that they launched in 2020. Onyx has managed $700 billion in transactions as of April 2023 .
JPMorgan's infrastructure investment in this area of India was not the first. Narendra Modi, Prime Minister since 2014, offers companies that set up within the International Financial Services Center of Gujarat International Finance Tec-City (better known as GIFT City) an 100% tax holiday for ten years. In GIFT City, bankers hold over $33 billion , with exemptions from tax rules that frustrate businesses in India, making it an attractive option for international investors.?
The project launches in early June 2023, with JPMorgan monitoring its process and banks' experiences with the technology. GIFT City hopes that the innovation of this new system will turn Gujarat into an alternative trading center like Singapore and Dubai.?
While this investment benefits the region, strategists connect JPMorgan's expansion into international markets and the emerging global de-dollarization . Brazil, Russia, India, China, and South Africa are all moving away from the dollar's dominance, causing the currency to lose its stronghold on global economies.
This news signals how instrumental blockchain technologies can become in the development of financial ecosystems. Equally, it represents the first of many movements of banks and governments to prepare for the possible declining impact of the US dollar.?
Additional Sources
ROUNDTABLE - GEN AI - ANALYSIS
Impact Of Generative AI In Tech Services Industry
July 2023: Generative Artificial Intelligence (GenAI) refers to algorithms that can generate new content. While OpenAI's ChatGPT is perhaps the most recognized example, many companies are developing GenAI in various contexts, such as voice, video, images, and code.
GenAI opens up new possibilities for rethinking the way we work. However, deep industry knowledge and proactive risk management are essential to harness its potential. Like Analytical AI, GenAI requires a robust data foundation, including flexible data infrastructure, cloud access, and proficiency in ML Ops. The successful implementation of GenAI will also necessitate new talent acquisition and effective change management.
We're on the cusp of a significant shift in the business landscape ushered in by GenAI. Just as the Computing era (starting in 1977) was succeeded by the Internet era (from 1995) and the Mobile era (from 2006), we believe 2023 will kickstart the AI "everywhere" era. This new era will be marked by the ability to generate near-human-level content at virtually no additional cost. It's predicted that GenAI will be a major catalyst for future labor productivity growth in developed economies. GenAI can significantly speed up about 50% of tasks without compromising quality.
Stay tuned for our next issue, where we'll dive deeper into our recent roundtable discussion. We'll be unpacking thoughts on the transformative impact of Generative AI on the tech services industry and exploring what this means for the future. Plus, we'll bring you the latest industry news, backed by data, to keep you at the forefront of the tech landscape. It's an issue packed with insights you’ll want to take advantage of!
JUST IN CASE YOU MISSED IT
DataOps.live gets $17.5M from Snowflake , others for DevOps-style tools aimed at data experts.
Autonomous Data Warehouse – could be a new Oracle.
Snowflake Launches Government & Education Data Cloud to enable data-informed government and enhance mission outcomes for citizens and students.
Palantir Launches Foundry for Manufacturing on AWS to the broader market.
Teradata deepens Dataiku integration to accelerate enterprise AI projects.
According to analysts, AMD's announcement represents the strongest challenge to Nvidia. The MI300X chip will start shipping to customers later this year.
In collaboration with consulting firms PwC and EY, Microsoft has taken steps toward ensuring the responsible and ethical use of AI technologies.
The recently announced cloud data and analytics platform, Microsoft Fabric, will directly compete with cloud giants Amazon and Google, offering more features and better data management.
INDUSTRY DATA
CIO Survey Data
The Recognize CIO Survey series is a regular panel of 250-500 CIOs in the U.S., depending on the time of year.
We use this data to track spending intentions, changes in technology, product preferences, strategic priorities, and talent challenges.
Which Gen AI products are your organizations using most?
Regarding leveraging Gen AI products, our survey found that organizations using Gen AI products (82%) use ChatGPT most frequently. This is followed by Copy.ai at 29%, with open-source solutions close behind at 28%. Bard and Claude are utilized by 27% and 25% of respondents, respectively, while Jasper is used by 24%. Chatsonic, however, appears less frequently used, with just 16% of organizations implementing it.
How has your organization approached Generative AI?
When asked about their approach to Generative AI technologies such as ChatGPT and Bard, 43% of organizations report having already deployed them in an enterprise application or process. Meanwhile, 21% of respondents have individuals experimenting on their own. Prototyping for enterprise use and large projects underway account for 17% and 13%, respectively. Interestingly, only 4% of organizations are not using Generative AI at all, and only 1% have explicitly banned its use.
What are your main concerns about accelerating the deployment of Generative AI?
When asked about the main concerns related to accelerating the deployment of Generative AI, security tops the list, with 52% of organizations expressing this concern. Complexity is the second primary concern at 39%. The need for hardware resources and the potential for inaccurate results garnered 33% of responses. The impact on jobs and sourcing talent to manage these systems is 31%. Lower return on investment is a worry for 20% of respondents. Interestingly, 8% express no material concerns or feel all concerns are manageable.
What automated code generation tool are your developers using?
As for the usage of automated code generation tools, GPT Code Clippy is the leader, employed by 50% of developers. GitHub Copilots closely follow at 45%. Bard is used by 26% of developers, while CodeComplete and Code Whisperer are used by 21% and 19% of developers, respectively. SourceGraphy Codey and Replit Ghostwriter are adopted by 17% and 14%, whereas Tabnine is at the lower end with 12%. Interestingly, 8% of respondents indicated that their developers do not use an automated code generation tool.
Will you build your own proprietary LLM that uses your own data?
Our survey reveals that many organizations are contemplating developing their own proprietary large language models. A significant portion of respondents (42%) confirmed they plan to do so, while another 41% are still evaluating or considering this possibility. Only a fraction (7%) indicated they have yet to make plans to build their own proprietary large language model.
How do you expect AI to impact your organization over the next two years.
Respondents hold largely positive views when asked about the expected impact of AI on their organization over the next two years. The majority (46%) predict AI will lead to significant use cases that drive productivity. 24% foresee select use cases, albeit without a major impact. A significant portion (22%) expect a transformational impact from AI. Conversely, a small fraction of organizations (7%) do not anticipate much impact from AI.