Why PostgreSQL for Generative AI?

Bala Narasimhan, Google’s Group Product Manager, explains why PostgreSQL is particularly well-suited for generative AI projects. “PostgreSQL is not just a powerful SQL database but also highly adaptable,” he states. The open-source nature of PostgreSQL ensures that it remains innovative, with continuous contributions from a global community. This flexibility is crucial for developers looking to implement advanced AI without getting locked into proprietary systems.

Shambhu Hegde, who focuses on making Cloud SQL an optimal platform for developers, highlighted the ease of transforming PostgreSQL into a vector database, essential for managing the complex data structures used in AI. “By installing a simple extension like PG vector, PostgreSQL can be immediately leveraged for generative AI applications,” Hegde notes.

Vector Databases and Generative AI

The conversation also centered on the benefits of using vector databases in generative AI applications. Vector databases excel in handling high-dimensional data vectors typical in AI, such as those generated in natural language processing or image recognition tasks. “When you choose Cloud SQL for PostgreSQL and convert it into a vector database, you retain all the operational database benefits while gaining the ability to perform sophisticated AI data operations,” Narasimhan adds. This approach negates the need for data migration to specialized databases, reducing complexity and potential data integrity issues.

Getting Started Fast with Jumpstart Solutions

Hegde recognizes the need for speed in today’s competitive tech landscape and introduces the Jumpstart solution. This toolkit on GitHub enables developers to launch a generative AI application in under 30 minutes. “It’s designed for rapid deployment, allowing developers to iterate on the fly without getting bogged down in lengthy development cycles,” he explains. This is particularly advantageous for developers who may not have deep AI expertise but are eager to explore generative AI capabilities.

Best Practices for Scalable Application Development

Deploying applications into production requires careful consideration of performance and scalability. Hegde stresses the importance of observability within PostgreSQL when deploying generative AI applications. Cloud SQL for PostgreSQL offers tailored observability features that help monitor query performance and system health. “Developers can utilize tools like Query Insights to examine detailed query execution plans and optimize performance based on real-time insights,” says Hegde.

Narasimhan adds, “With Cloud SQL for PostgreSQL, you get the scalability and reliability of Google Cloud, ensuring that your applications are both powerful and secure.”

PostgreSQL: A Compelling Combination

PostgreSQL offers a compelling combination of flexibility, power, and community support for developers looking to harness the full potential of generative AI. Whether developing complex AI algorithms or needing a reliable database solution for enterprise applications, PostgreSQL integrated with Google Cloud’s SQL services provides a robust foundation to innovate and scale. As AI evolves, tools like PostgreSQL will remain indispensable for developers aiming to push the boundaries of what’s possible.

In a recent interview with CNBC’s “Mad Money,” Anil Chakravarthy, CEO of Informatica, shed light on the company’s groundbreaking work and its profound impact on the technology landscape. Chakravarthy emphasized Informatica’s critical role in empowering organizations to harness the power of data. He stated, “We spend a ton of time talking about the rise of cloud computing and Big Data because these are long-term stores. People don’t understand how long and far out they can go.” Indeed, with the exponential growth of connected devices, the volume of data being generated is staggering, making effective data management more crucial than ever before.

Chakravarthy highlighted Informatica’s impressive track record, noting that the company serves over 10,000 customers, including top Fortune 100 companies and a significant portion of the Forbes Global 2000. Despite being privately held, Informatica has experienced remarkable growth, with last year’s revenue reaching an impressive $1.2 billion. “We’re not public yet, but among the companies that are private, we’re the fastest grower,” Chakravarthy remarked.

At the heart of Informatica’s success lies its ability to empower organizations to unlock the value of their data. Chakravarthy explained, “This space around data management… it’s powering a lot of these new generations of AI-powered intelligence.” By providing a comprehensive platform for managing and analyzing data, Informatica enables companies to derive actionable insights and gain a competitive edge in today’s data-driven economy.

One of the key areas where Informatica’s technology is making a significant impact is in the life sciences industry. Chakravarthy cited examples of how Informatica’s solutions are helping companies like Amgen and Johnson & Johnson manage and analyze patient data, clinical trials, and drug compounds, accelerating drug discovery and improving patient outcomes.

Moreover, with the recent acquisition of AI-powered customer insights company AllSight, Informatica is poised to enhance further its capabilities in delivering actionable insights across industries. Chakravarthy announced, “We just announced the acquisition of a little company called AllSight, the leader in AI-powered customer insights. That’s the kind of platform that’s driving all of these next-generation uses of data in companies like life sciences.”

Informatica’s strategic partnerships with industry leaders like Salesforce, Adobe, and Microsoft highlight its central role in the technology ecosystem. Chakravarthy emphasized the importance of these partnerships, stating, “We integrate really well with Salesforce, Adobe, Microsoft, etc., to make that happen.” By seamlessly integrating with these platforms, Informatica enables companies to leverage their data more effectively and drive innovation across their organizations.

Looking ahead, Informatica remains focused on driving innovation and expanding its market presence. As the demand for data management solutions grows, Informatica is well-positioned to capitalize on emerging opportunities and cement its status as a leader in enterprise cloud data management.

In conclusion, Informatica’s story is a testament to the transformative power of technology and the critical role that data management plays in driving business success. As companies increasingly rely on data to fuel their growth and innovation, Informatica’s innovative solutions will continue to play a vital role in shaping the future of enterprise technology.

Mohan’s remarks addressed inquiries about the data sources utilized in AI model training, particularly concerning concerns surrounding OpenAI’s recent AI model, Sora, and its potential use of YouTube data. Mohan stressed YouTube’s commitment to transparency and adherence to precise terms of service, emphasizing the platform’s respect for creators’ intellectual property rights.

“At YouTube, we understand the importance of data privacy and respect the rights of our creators,” Mohan stated. “Strict terms of service and contractual agreements govern any utilization of creator content in AI development.”

The conversation delved deeper into the specifics of YouTube’s data policies and the safeguards to protect user privacy and intellectual property. Mohan highlighted the platform’s rigorous approach to data usage, ensuring that AI development aligns with ethical standards and legal obligations.

“While enhancing search functionality through data scraping is essential, we are unequivocal in our stance against unauthorized downloading of transcripts or video segments,” Mohan explained. “Our terms of service explicitly prohibit such actions, reflecting our commitment to upholding creator rights.”

Moreover, Mohan emphasized YouTube’s efforts to support creators and ensure fair compensation for their work. “We are committed to fostering a vibrant ecosystem conducive to innovation and creativity,” he remarked. “This includes providing creators with tools and resources to protect their intellectual property and maximize their earning potential.”

As the debate surrounding data ethics and AI intensifies, YouTube faces growing scrutiny over its data practices and their implications for user privacy and intellectual property rights. Mohan’s remarks underscore the need for greater transparency and accountability in AI development as platforms navigate the complex intersection of technology, ethics, and user rights.

Moving forward, YouTube remains steadfast in its commitment to upholding its core mission of empowering creators and delivering compelling user experiences while adhering to the highest standards of data ethics and privacy protection. Mohan’s interview revelations highlight the ongoing challenges and opportunities in responsibly leveraging data to drive innovation and progress in the digital age.

Rifi’s platform, housed in the cloud, empowers data teams to stay ahead of the curve by offering unparalleled visibility into their operations. Sanjay highlights two key features that define Rifi’s offering. Firstly, the platform helps teams manage their budgets effectively, ensuring they don’t exceed allocations for cloud services like Snowflake or BigQuery. This proactive approach to cost management saves money and fosters trust within organizations, as data flows smoothly and reliably.

Secondly, Rifi prioritizes the time of data teams, recognizing that efficiency is crucial for building trust and making informed decisions. Sanjay notes that Rifi has enabled some customers to drastically reduce escalations from data teams to their businesses, a testament to the platform’s ability to streamline operations and increase productivity.

One of Rifi’s standout success stories involves a public company with a $10 billion market cap. Within just three weeks of implementing Rifi’s solution on BigQuery, the company noticed a significant increase in failed jobs, indicating issues with data accessibility. Instead of resorting to the traditional approach of requesting more resources, Rifi’s platform enabled the company to identify the root cause of the problem quickly: certain user and query patterns consuming excessive capacity. By addressing these issues promptly, the company was able to free up nearly a quarter-million dollars worth of capacity, demonstrating the tangible impact of Rifi’s technology on the bottom line.

When asked about Rifi’s approach to innovation, Sanjay emphasizes the company’s commitment to listening to its customers. With clients spanning various industries, including public, healthcare, finance, and startups, Rifi understands the diverse needs and challenges facing data teams today. By staying attuned to customer feedback and continuously iterating on its platform, Rifi ensures that it remains at the forefront of innovation, delivering solutions that meet the dynamic demands of the modern tech landscape.

In conclusion, Rifi’s groundbreaking approach to data observability and rapid issue detection is revolutionizing the way organizations harness and leverage their data. By combining cutting-edge technology with a customer-centric approach, Rifi empowers data teams to navigate the complexities of today’s digital world with confidence and agility.

Data breaches and ransomware attacks continue to increase, with organizations of all sizes being impacted. In all cases, the sooner consumers are notified, the sooner they can take action to protect their data or be on guard against phishing attempts.

The FCC wants to ensure telco consumers have as much notice as possible, with new rules limiting how long companies have to disclose a breach impacting consumer data. The agency outlined its the changes in its final rule on the matter:

In this document, the Federal Communications Commission (Commission) modifies the Commission’s data breach notification rules to better ensure that providers of telecommunications, interconnected Voice over Internet Protocol (VoIP), and telecommunications relay services (TRS) are held accountable in their obligations to safeguard sensitive customer information, and to provide customers with the tools needed to protect themselves in the event that their data is compromised.

The agency limits the new rule to instances where a company has reason to believe consumer harm is likely to occur.

The Commission also adopts its proposal to require carriers to notify the Commission, in addition to the Secret Service and FBI, as soon as practicable, but no later than seven business days, after reasonable determination of a breach. The Commission next eliminates the requirement that carriers notify customers of a breach in cases where a carrier can reasonably determine that no harm to customers is reasonably likely to occur as a result of the breach.

The new rule will hopefully help consumers act sooner to protect their data in the wake of a breach.

According to new research by VPN provider SurfShark, the US government makes the most requests for user data from Big Tech companies than any other jurisdiction in the world. The company analyzed data requests to Apple, Google, Meta, and Microsoft by “government agencies of 177 countries between 2013 and 2021.”

The US came in first with 2,451,077 account requests, more than four times the number of Germany, the number two country on the list. In fact, the US made more requests than all of Europe, including the UK, which collectively came in under 2 million.

While the US and EU were responsible for a combined total of 60% of all data requests, the US “made 8 times more requests than the global average (87.9/100k).”

The number of accounts being accessed is also growing, with a five-times increase in requests from 2013 to 2021. The US alone saw a 348% increase during the time frame, and the scope and purpose of the requests are expanding.

“Besides requesting data from technology companies, authorities are now exploring more ways to monitor and tackle crime through online services. For instance, the EU is considering a regulation that would require internet service providers to detect, report, and remove abuse-related content,” says Gabriele Kaveckyte, Privacy Counsel at Surfshark. “On one hand, introducing such new measures could help solve serious criminal cases, but civil society organizations expressed their concerns of encouraging surveillance techniques which may later be used, for example, to track down political rivals.”

The report also sheds light on which companies comply the most versus which ones push back against requests. For all of its privacy-oriented marketing — “what happens on your iPhone stays on your iPhone” — Apple complies with data requests more than any other company, handing it over 82% of the time.

In contrast, Meta complies 72% of the time, and Google does 71% of the time. Microsoft, on the other hand, pushes back the most among Big Tech companies, only handing data over 68% of the time.

The findings may also put a dent in US efforts to ban TikTok and other foreign apps under the guise of protecting user privacy and data.

The global big data and analytics market is expanding every year, currently sitting just below $300 billion USD. Yet, just having access to the data architecture that enables you to collect and process data doesn’t mean that a business does so in an efficient or compliant way. Data compliance is one of the leading issues in modern business, with the fast scaling required by some businesses leading them to bad data management habits.

In order to store and utilize huge quantities of data, businesses need to take an active and proactive approach to data compliance. In this article, we’ll dive into the core issues that businesses encounter with compliance and point toward leading strategies to fix them. 

Let’s dive right in.

What Is Data Compliance?

Data compliance is an intersectional field that balances between operational data use and legal requirements. When using data, especially for analytics or business operations, it’s important to establish clear guidelines about how you are using that data and how you protect it. Data privacy, availability, and integrity are all considerations in data compliance.

Another pillar of compliance that is important to recognize is that, depending on the region that your business operates from, you may also have regulatory compliance issues to overcome. For example, Europe has strict privacy laws when it comes to handling customer data, which you must comply with or face fines.

Data compliance protects both your company and your customers, keeping all of your data, your customer data, and your private records safe. As the rising cyber threat continues to focus on the enterprise sector, expert control of data compliance and privacy is more needed than ever before.

How To Streamline Data Compliance

Data compliance may seem staggeringly complex at first glance, but is really only a representation of the very best laws, ideas, and data protection strategies to keep your business safe. As a deeply methodical process, it’s a good idea to get started with data compliance from day one.

There are a number of ways you can streamline data compliance in your organization.

Establish a Workflow

The single biggest factor that will lead to breaches in data compliance is your own employees. If they don’t know how to stay on the side of compliance, you’ve probably not made your processes obvious enough for them.

Be sure to establish a comprehensive workflow of how you should handle data when it enters your business. Create detailed guides about the use and distribution of data that you then teach to all new employees during their onboarding.

Every staff member should understand the importance of data compliance, what it means, and how to use data in a compliant manner.

Audit, Audit, Audit

Even for businesses that already have comprehensive data compliance workflows and frameworks set out, it’s a good idea to audit your business frequently. Not only does a regular audit ensure that all of your processes are being upheld and followed to the correct standard, but it also helps to create a permanent record of your work.

Audit Trails are required by many regulatory bodies, with proof of your regular system audits and the systems logged in the process acting as evidence in this regard. Depending on your geographical location, the specific audit laws you have to work to may vary. Some states in the USA are much more strict than others.

To air on the side of caution, we recommend that you audit your internal processes at least once a year. However, if you’ve experienced a data breach or are incorporating new data technologies, then you should conduct an audit much more frequently. 

Automate Where Possible

Once you have a data governance scheme in place, you’re now in a position to automate much of the initial contact points with data. There are numerous data compliance automation tools and strategies that you can put to use. Not only do these save tremendous amounts of time, especially those that focus on granular data monitoring, but they will also save your business capital in the long run.

There are several areas where you can use automation to streamline the data compliance process:

• Incident Management – Around 40% of all businesses will experience a data disaster event at least once a year. This staggeringly high figure reminds us that creating an incident management system will allow you to respond to any events much more rapidly.

• Recovery and Backup – One core area that you can optimize with automation is any recovery and backup processes. When creating a catalog of historical data, it’s always a good idea to create failsafe backup systems. If needed, you can use these backups to resort to an earlier version of your data systems before a compliance breach occured.

• Schema and Management – Once you have a specific schema of data that you want to collect, you can then create safeguards that flag any data that does not strictly follow this format. Any breaches or inconsistencies will alert a developer, allowing them to take a look at any strange data instead of monitoring every single dataset that you ingest.

The ability to successfully automate aspects of data compliance will only grow in the coming years. With advancements in AI, ML, and other emerging technologies, we’ll be able to spend more time focusing on the mission-critical aspects of compliance and less on the day-to-day monotonous tasks.

Final Thoughts

The importance of the use of data cannot be overstated. As a leading tool that guides us toward better strategies, more effective use of resources, and more precise future planning, data is essential in modern business. Yet, without a comprehensive understanding of data compliance, businesses are unable to make the most of the data they have at their disposal.

By investigating your current data practices, including how you store and interact with data in your cloud data warehouse, businesses will be ready to take a more effective position on data compliance. By utilizing the strategies, examples, and systems we’ve suggested in this article, you’ll be able to construct a rigorous compliance program that secures your data while passing regulations.

Oracle has long been the dominant player in the database market. Even as the cloud has grown in importance, Oracle has managed to carve out a meaningful share of the market, thanks in large part to the strength of its database platform. Many customers see it as a full turn-key solution, combining the cloud and database solutions necessary. In spite of that, according to a report by Bloomberg, Oracle’s database dominance may be under threat from cloud-first rivals.

Bloomberg cites the example of Shutterfly, which recently made the decision to move its database to the cloud. Despite relying on Oracle for years, the company decided to go in a different direction with the transition.

Read more: Larry Ellison Touts Oracle Cloud’s Reliability in Wake of AWS Outage

“The amount of time and energy that was consumed purely running just the plumbing was immense,” Chief Technology Officer Moudy Elbayadi said in an interview. A review of the existing options on the market led Shutterfly to conclude that Oracle’s solutions didn’t “fit our desires to have that level of openness and flexibility,” Elbayadi added.

Unfortunately for Oracle, Shutterfly isn’t an isolated example. JPMorgan, Nasdaq Inc, JetBlue Airways Corp, and Automatic Data Process Inc are among the list of companies transitioning to non-Oracle options.

“We have actually quite rapidly been reducing our Oracle footprint,” said Nikolai Larbalestier, Nasdaq’s senior vice president of cloud strategy and enterprise architecture. “There are plenty of good alternatives today.”

Part of the problem stems from the complexity involved in running Oracle’s database and the cost to the client company of doing so. Mythical Games CEO John Linden emphasized the issue, despite his firm being valued at $1.2 billion.

“Oracle hits us up every week,” he said. But “we’d have to have a massive team in place to run it appropriately.”

See also: Google, Microsoft, and Oracle Had the Most Vulnerabilities in Early 2021

Just as significant, Oracle’s tools seem to be developing a reputation for not being up to par with the latest developments, making the prospect of working with them unappealing to many developers.

“I can’t even hire people if I told them that we majorly use Oracle,” Yao Morin, chief data officer at JLL Technologies, told Bloomberg. “People are yearning for better tools.”

To be clear, Oracle is still the company to beat in the database market, especially among companies that want on-premise database solutions. Nonetheless, the company clearly has some significant areas it needs to improve on if it wants to remain relevant in the coming years. Otherwise, it may find itself in the same situation as IBM when personal computers replaced mainframes — the undisputed leader of a niche market.

Microsoft and Oracle are two of the leading cloud providers, although both companies lag behind market leader AWS. The two companies are teaming up to give customers the ability to run OCI on Azure servers.

“We have a real opportunity to help organizations bring their mission-critical applications to the cloud so they can transform every part of their business with this next generation of AI,” said Satya Nadella, Chairman and CEO, Microsoft. “Our expanded partnership with Oracle will make Microsoft Azure the only other cloud provider to run Oracle’s database services and help our customers unlock a new wave of cloud-powered innovation.”

“Most customers already use multiple clouds,” said Larry Ellison, Oracle Chairman and CTO. “Microsoft and Oracle have been working together to make it easy for those customers to seamlessly connect Azure Services with the very latest Oracle Database technology. By collocating Oracle Exadata hardware in Azure datacenters, customers will experience the best possible database and network performance. We are proud to partner with Microsoft to deliver this best-in-class capability to customers.”

The partnership is a win for both companies. Oracle, the king of database products, has lost customers who have migrated to competitors’ cloud offerings. By partnering with the second-largest cloud provider, Oracle offers customers a viable cloud migration path that keeps them using the company’s database products.

At the same time, Microsoft benefits by being the only other cloud provider to offer Oracle Cloud Infrastructure Database Services, setting it up to be a preferred option for Oracle customers looking to migrate to a more widely used cloud platform.

The significance of the partnership was not lost on the two tech leaders.

“It’s actually my first time in Redmond. It’s hard to believe,” Ellison said. “I waited until very late in my career to make the trip.”

“It took us 45 years, but here we are!” Nadella added.

The FTC banned X-Mode, and its successor Outlogic, from selling precise location data in early January. The agency has now followed suit with a similar ban on InMarket Media. The ban prohibits the company from selling or licensing precise location data in response to the company’s failure to provide the necessary disclosures or obtain the necessary consent from consumers.

“All too often, Americans are tracked by serial data hoarders that endlessly vacuum up and use personal information. Today’s FTC action makes clear that firms do not have free license to monetize data tracking people’s precise location,” said FTC Chair Lina M. Khan. “We’ll continue to use all our tools to protect Americans from unchecked corporate surveillance.”

The two cases should service as a warning to other data brokers not to play fast and loose with consumers’ privacy.

Oracle bet big on health, purchasing medical records giant Cerner for $28 billion last year. According to Business Insider, Oracle Health laid off hundreds of employees Thursday, rescinded existing job offers, and eliminated some open positions.

Insider’s sources say laid-off employees will receive four weeks severance pay, as well as an additional week for every week of employment. Vacation days will also be paid out.

The impetus for the layoffs appears to be issues with Cerner’s contract with the US Department of Veterans Affairs, where unexpected issues have hampered rollout of the company’s platform. The issues were severe enough to trigger a contract renegotiation, one that will hold Cerner more accountable for issues and that led to less work for the company than previously anticipated.

In its latest quarterly results (via CNBC), Oracle reported revenue of $13.84 billion, beating analysts’ expectations of $13.74 billion. Similarly, earnings came in at $1.67 per share, beating expectations of $1.58 per share.

As CNBC points out, Oracle’s results are helping drive the stocks to new highs, with it “up 73% over the past 12 months.” As result, with the exception of Nvidia, Oracle’s stock has outperformed “all large-cap enterprise tech stocks” for the same period. The stock’s current price also makes 2023 the best year for Oracle investors since the dot-com boom.

The company’s current status is surprising to some who had written Oracle off as a bit player in the cloud market and felt the company was losing ground on too many fronts.

Much of the credit is going to founder Larry Ellison, who has helped the company make adjustments and catch up in critical markets.

“Let’s give him credit where it’s finally due,” said Eric Lynch, managing director of Scharf Investments. “The upside case is finally coming through.”

The company’s partnership with AI firm Cohere is also paying off, especially given Cohere’s focus on the enterprise. This makes it an ideal match for Oracle.

Ellison has been saying for years that Oracle had an advantage over its competitors, especially in the realm of integration, stability, and reliability.

“Let me close with a note that I’m going to paraphrase from a very large telecommunications company who uses our cloud and all the other three North American clouds — Google, Amazon and Microsoft,” Ellison said in late 2021. “And the note basically said the one thing we’ve noticed about Oracle, Oracle’s cloud, is that it never ever goes down. We can’t say that about any of the other clouds. We think this is a critical differentiator.”

It appears the industry is finally taking note of what Ellison has been saying all along.

States With the Most Data Breaches

On an individual state basis, California, Oregon, Maryland, Georgia, and Virginia have suffered the most losses. However, nearly every state across the U.S. has suffered losses in the millions, some even in the billions. In an attempt to tackle this problem, experts have identified the major causes of data loss in the United States. One reason is human error, as many companies or organizations suffer from employees that are prone to accidental deletion or misclicks, or simply have a lack of training. Another cause is malware. Phishing is far too common in the online space, as is spoofing and ransomware. The final main cause is unexpected events, such as hardware failure, software glitches, or external natural disasters.

Looking at Data Breaches on a Global Scale

Although countries around the globe, much like the United States, suffer from data loss, countries aren’t the only entities that are affected. In fact, 45% of retailers have reported an increased size, severity, and scope of cyber attacks against their data. In addition, between 2021 and 2022, over 5,000 global businesses experienced and confirmed data loss within their company. This is a widespread issue, as important industries like finance, healthcare, public administration, manufacturing, and transportation are all heavily affected by this crisis. Some examples of these data breaches are the finance incident in Ukraine or the transportation incident in Japan. In Ukraine in 2018, 100GB of data was exfiltrated from a loan services company. In 2022 in Japan, Toyota lost 300,000 customers’ emails to hackers.

Data loss is not a new concept to our global society, as we have destroyed or lost plenty of historical data over the years. Perhaps the most famous and devastating data breach in history was the burning of the Library of Alexandria, losing an estimated 571.4GB of data in one fell swoop. Other examples include, but are not limited to, the destruction of the Royal Library of Ashurbanipal or the loss of the Maya Religious Codices. 

Conclusion

It is clear that data loss remains both a global and historical certainty, regardless of whether that loss is physical or digital. There is no way of ensuring that data is never lost, leaked, or stolen, but it is beneficial to be prepared for the possibility of these tragedies. In a world that is largely based online and the value of digital assets increases, the risk for these cybersecurity breaches also increases. Whether it is personal data or company data, cyber security affects us all, and is an important factor to consider when moving forward with data storing and sharing.

Microsoft is rolling out ChatGPT across multiple products and has no objection to its own employees using the tech. However, the company wants to make sure no sensitive information is shared with the AI.

“Please don’t send sensitive data to an OpenAI endpoint, as they may use it for training future models,” a senior engineer wrote in an internal post that was reviewed by Business Insider.

The memo demonstrates one of the biggest challenges moving forward with large language model AIs, namely controlling what information it has access to, and how that information will be used if it is shared.

ChatGPT is a conversational AI that learns from its interactions and what people type into it. As such, it’s not surprising that Microsoft wants to make sure no sensitive information is shared with it, since the AI could then end up using that information in its responses to users.

“Human beings sign NDAs and consequently have incentives to be careful in how they share information. But large language models such as ChatGPT do not have the ability to reason about such issues, at least by default,” Vincent Conitzer, Carnegie Mellon University computer science professor and director of its AI lab, told Insider.

Microsoft’s caution is one other companies would do well to imitate.

On January 11, a Notice to Air Missions (NOTAM) system outage impacted some 11,000 flights in the US. In the days after the incident, the FAA issued a statement explaining the cause:

A preliminary FAA review of last week’s outage of the Notice to Air Missions (NOTAM) system determined that contract personnel unintentionally deleted files while working to correct synchronization between the live primary database and a backup database. The agency has so far found no evidence of a cyber-attack or malicious intent. The FAA continues to investigate the circumstances surrounding the outage.

The FAA made the necessary repairs to the system and has taken steps to make the NOTAM system more resilient. The agency is acting quickly to adopt any other lessons learned in our efforts to ensure the continuing robustness of the nation’s air traffic control system.

According to Reuters, the FAA has implemented “a one-hour synchronization delay for one of the backup databases. This action will prevent data errors from immediately reaching that backup database.” The FAA also said it “now requires at least two individuals to be present during the maintenance of the (messaging) system, including one federal manager.”

Google Cloud is already a popular option among startups and developers. The company is expanding its partnership with MongoDB to provide integrated database and data services.

As partners, Google Cloud and MongoDB co-engineer streamlined integrations between MongoDB Atlas and many Google Cloud services to make it easier to deploy apps (Dataflow, GKE, Cloud Run), pull in data from other sources (Apigee), run in flexible multi cloud environments (Anthos), easy deployment of MEAN stack, and Terraform and analyze data (BigQuery, Vertex AI).

Startups will benefit from Google Cloud’s global reach, giving them the ability to expand and scale as needed.

Signups will also receive significant savings on Google Cloud and Firebase:

If you’re early in your startup journey and not yet backed with equity funding, you’ll have access to $2,000 of Google Cloud credits. If you are, your first year of Cloud and Firebase usage is covered with credits up to $100,000. Plus, in year two get 20% of Google Cloud and Firebase usage covered, up to an additional $100,000 in credits.

Similarly, signups will receive free credits for MongoDB:

Free credits for MongoDB Atlas, including usage of the core Atlas Database, in addition to extended data services for full-text search, data visualization, real-time analytics, building event-driven applications and more to supercharge your data infrastructure

The expanded partnership looks to be a big win for startups.

Dan Springer, CEO of DocuSign, discusses how the COVID-19 pandemic has accelerated digital transformation and he says that businesses are not going back to a manual world:

COVID Pandemic Accelerated Digital Transformation

We’ve been really pleased with the growth we’ve had since going public a few years. We have also seen significant acceleration since the COVID-19 pandemic. It’s obviously a horrible pandemic and our number one priority has been the health and wellbeing of our employees so we can take good care of our customers. As you can see in our Q1 earnings we did see an acceleration of our bookings to 59 percent.

Traditionally, if you look at the billings-type metric they have been in the mid-30s’. A significant portion of that (increase) was due to increased use cases from customers driving that digital transformation faster with services like DocuSign.

Companies To Stay In This Digitally Transformed World

One of the things we’ve seen with the pandemic impact is that it has really accelerated the path that companies were already on to drive that digital transformation. We don’t see companies after the pandemic settles down going back and saying they want more paper and more manual processes.

Once they’ve got the benefits from that efficiency in their business, the better customer experience, and the better employee experience, they’re going to stay in a digitally transformed world. They are going to use DocuSign and other fantastic services to do that.

The Future Is Going To Have eSignature At The Center

We really think that the future is going to have eSignature at the center of what we call the overall Agreement Cloud. Companies want to be more agreeable. They want to be easier to do business with and be easier to do business for. They’re going to not just use DocuSign for signature but all of the other components of preparing agreements and managing those agreements digitally once they’ve been created. That’s why we’re excited about our very robust future.

We just past a billion dollars in revenue (for DocuSign eSignature). We are only four percent penetrated today and we’re six times larger than the next biggest player in the space. There’s not a lot of penetration yet in that core business. Notary is still predominantly done manually. We are making investments there. We believe we can bring the same ease of use that we brought to eSignature we can bring to notary.

AI To Power The DocuSign Agreement Cloud

Much bigger than that, even expanding upon the opportunity of eSignature is that broader Agreement Cloud opportunity. We think this is the next big cloud opportunity. You are going to see companies increasingly say I don’t just want to do the workflow and signature. I also want to drive the creations of those agreements. I want to think about artificial intelligence and search capability to manage my agreements. This would enable me to actually manage my business and make my company more agreeable.

Those are some of the investments we’re making. That’s why we just finished the acquisition of Seal Software last month so we can bring additional artificial intelligence and analytic capability to help people run their businesses better.

Microsoft operates the second-largest cloud platform and, as such, operates data centers around the world. As one of the world’s largest growth markets, Asia represents tremendous opportunity for the company.

“Absolutely. We’re very, very bullish about what’s happening in Asia,” Satya Nadella, said in an interview with CNBC’s Tanvir Gill.

Nadella singled out two countries as especially important to the company’s future: China and India.

“We’re absolutely committed to all of these countries and in China too,” Nadella said. “Today, we primarily work to support multinational companies that operate in China and multinational companies out of China.”

Similarly, while India is important to the company’s future, Microsoft sees significant changes to the market.

“Microsoft’s presence in India was about mostly multinational companies operating in India. But for now, it’s completely changed,” he said.

“It’s the reverse where these companies who are innovating in India, whether it’s the big large conglomerates, or the new startups, are all using [artificial intelligence] cloud technology to be able to innovate and create services that are obviously popular in India and elsewhere,” he added.

In all, Nadella said Microsoft plans on investing in at least 11 different regions.

Data scraping is the process of using automated methods and scripts to collect data from a website. The data may be publicly available or require access. News of the scraping breach first broke in early 2021, although the actual incident occurred prior to 2020. In all, some 533,000,000 Facebook accounts were impacted.

Ireland’s Data Protection Commissioner (DPC) has now levied the third-largest fine against Meta, saying the company did not do enough to protect its users’ data and prevent personal information, phone numbers, email addresses, and more from being scraped.

According to Independent.ie, some 1.3 million Irish Facebook accounts were impacted. Some of the impacted accounts included “gardai, sitting judges, prison officers, social workers, journalists and others.” The breach also coincides with a spike in scam attempts across the EU and Ireland.

“The material issues in this inquiry concerned questions of compliance with the GDPR obligation for data protection by design and default,” the DPC said in a statement. “The DPC examined the implementation of technical and organisational measures pursuant to Article 25 [of] GDPR.”

The investigation was evidently started last year, after news of the breach.

“The DPC commenced this inquiry on 14 April 2021, on foot of media reports into the discovery of a collated dataset of Facebook personal data that had been made available on the internet,” the DPC statement said.

“The scope of the inquiry concerned an examination and assessment of Facebook Search, Facebook Messenger Contact Importer and Instagram Contact Importer tools in relation to processing carried out by Meta Platforms Ireland Limited during the period between 25 May 2018 and September 2019.”

To make matters worse, Facebook apparently is not interested in accepting full responsibility for the incident or fully committing to preventing such incidents in the future. In fact, as we previously covered at WPN, Facebook accidentally sent a memo to a journalist in which the company complained about the negative coverage it was receiving over the breach.

In the memo, the company also outlined its goals moving forward, including efforts to “normalize the fact that this activity happens regularly.”

Thankfully, Ireland’s DPC doesn’t believe data scrapping should be accepted as ‘normal’ and is holding Meta’s feet to the fire.

Small computer repair shops may be a common site, but a new report indicates customers should be wary before taking their computers to them. Researchers at University of Guelph in Ontario, Canada took laptops to 12 repair shops. The laptops were fully functional, except for a disabled audio driver. The researchers specifically chose that issue, since it is easy to diagnose and repair, and does not require access to personal files.

The researchers populated the computers with what appeared to be personal information, online accounts, a crypto wallet, and a variety of sexual and non-sexual pictures. The researchers also made it appear that half the computers belonged to men and half to women.

In 50% of cases, the researchers found that personal files were accessed by the repair shop, although unsurprisingly the computers that seemed to be belong to women were much more likely to have their data accessed. In at least two cases, one for a male customer and one for a female, data was copied and saved onto personal devices.

“We were blown away by the results,” Hassan Khan, one of the researchers, said in an interview with Ars Technica. The researchers were especially concerned with the data copying.

“We thought they would just look at [the data] at most,” Khan added.

With few if any real privacy safeguards in place, most customers would do well to take their computers to reputable large companies, at least until small shops get with the program, in terms of privacy.