Results

Microsoft uses Oracle Cloud Infrastructure for Bing conversational workloads

Microsoft uses Oracle Cloud Infrastructure for Bing conversational workloads

Microsoft is using Oracle Cloud Infrastructure for its Microsoft Bing generative AI searches.

Oracle announced the multi-year agreement with Microsoft in a press release. What we don't know is whether Microsoft is using Oracle Cloud to overflow Bing workloads or completely due to efficiency and/or procurement of Nvidia GPUs.

The two companies recently outlined a partnership. Oracle also fired up Nvidia-powered instances and apparently has been able to procure GPUs. Oracle and Microsoft have a history of partnership announcements that seem to be refreshed often. For instance, Oracle and Microsoft outlined an interoperability partnership between clouds in 2019. In 2022, the two companies announced the general availability of Oracle Database Service for Microsoft Azure.

Also: Oracle adds vector search to Oracle Database 23c, melds generative AI, transactional data | Oracle's Q1 better than expected and Ellison loves generative AI

Although Microsoft CEO Satya Nadella and Oracle CTO Larry Ellison may seem like odd bedfellows, both companies have joint customers and mutual rivals in AWS and Google Cloud.

The Oracle Cloud-Bing announcement could also simply be a headliner use case for enterprises. Microsoft is using Oracle Cloud along with its Azure AI infrastructure for inferencing for Bing and managed services such as Azure Kubernetes Service to orchestrate workloads. The connection also uses Oracle Interconnect for Microsoft Azure.

Constellation Research analyst Holger Mueller said that "generative AI has the potential to change the cloud market landscape." Time will tell if the Microsoft-Oracle partnership for Bing conversational workloads is a data point for cloud leadership changes.

Mueller made the following points about reading the tea leaves with the Oracle and Microsoft partnership. 

  • It is a milestone because no cloud vendor has ever moved internal workloads - or any workloads to a partner/competitor. 
  • It is a sign that Microsoft may be maxxed out on capacity. 
  • It is a sign it needs and does charge more to customers than it wants to afford internally.
  • Oracle gets the workload of the second largest search engine for generative AI search and evidently have the capacity.
  •  Oracle seems to get Nvidia chips at a better rate than Microsoft.

Research

Data to Decisions Digital Safety, Privacy & Cybersecurity Innovation & Product-led Growth Tech Optimization Future of Work Next-Generation Customer Experience Oracle SaaS PaaS IaaS Cloud Digital Transformation Disruptive Technology Enterprise IT Enterprise Acceleration Enterprise Software Next Gen Apps IoT Blockchain CRM ERP CCaaS UCaaS Collaboration Enterprise Service ML Machine Learning LLMs Agentic AI Generative AI Robotics AI Analytics Automation Quantum Computing Leadership VR Chief Information Officer Chief Technology Officer Chief Information Security Officer Chief Data Officer Chief Executive Officer Chief AI Officer Chief Analytics Officer Chief Product Officer

OpenAI launches GPTs as it courts developers, models for use cases

OpenAI launches GPTs as it courts developers, models for use cases

OpenAI launched new developer tools, models and GPTs designed for specific use cases.

On its first developer day, OpenAI moved to expand its ecosystem, enable developers, and leverage the popularity of its models so they can be customized.

For OpenAI, the set of announcements brings it closer to where enterprises are going--smaller large language models (LLMs) and generative AI that is tailored to tasks. These task-specific models--called GPTs--will roll out today to ChatGPT Plus and Enterprise users. Constellation Research analyst Holger Mueller said:

"Buried in the press release in a side sentence is what is the biggest challenge in enterprises adopting LLMs. Important information is in OLTP systems and can't be accessed by LLMs. If the OpenAI GPTs capability to access OlTP databases work - we will enter the next generative AI era." 

In a blog post, OpenAI said:

"Since launching ChatGPT people have been asking for ways to customize ChatGPT to fit specific ways that they use it. We launched Custom Instructions in July that let you set some preferences, but requests for more control kept coming. Many power users maintain a list of carefully crafted prompts and instruction sets, manually copying them into ChatGPT. GPTs now do all of that for you."

While OpenAI's ChatGPT is being used for context specific use cases in Microsoft productivity applications, the company is also looking to put its own mark on its models. The game plan for OpenAI is to build a community and launch a GPT Store, which will feature GPTs across a broad range of categories. Developers will get a cut of the proceeds from the GPT Store.

Key points about GPTs:

  • Your chats with GPTs are not shared with developers. If a GPT uses third party APIs you have control over what data can be sent to API and what can be used for training.
  • Developers can use plug-ins and connect to real-world data sources.
  • Enterprises can use internal-only GPTs with ChatGPT Enterprise. These GPTs can be customized for use cases, departments, proprietary data and business units. Amgen, Bain and Square are early customers.
  • OpenAI also launched Copyright Shield, which will indemnify customers across ChatGPT Enterprise and the developer platform.

That backdrop of GPTs complements a bevy of other OpenAI launches that are more aligned with where the LLM market is headed.

Related: Why generative AI workloads will be distributed locally | Software development becomes generative AI's flagship use case | Enterprises seeing savings, productivity gains from generative AI | Get ready for a parade of domain specific LLMs

Here's the breakdown.

  • OpenAI launched GPT-4 Turbo, which is up to date through April 2023 and has a 128k context window. It is also optimized for a 3x cheaper price for input tokens and 2x cheaper price for output tokens relative to GPT-4.
  • GPT-4 Turbo will also be able to generate captions and analyze real-world images in detail and read documents with figures.
  • DALL-E 3 has been updated to programmatically generate images and designs. Prices start at 4 cents per image generated.
  • The company improved its function calling features that enable you to describe functions of your app to external APIs to models. Developers can now call multiple functions, send one message to call multiple functions and improve accuracy.
  • GPT-4 Turbo supports OpenAI's new JSON mode.
  • OpenAI added reproducible outputs to its models for more consistent returns. Log probabilities will also be released so developers can improve features.
  • The company released the Assistants API in beta for developers to build agent-like experiences in applications. These assistants use Code Interpreter to write and run Python code, Retrieval to leverage knowledge outside of OpenAI models and Function calling.
  • Developers will also get text-to-speech APIs to generate human-quality speech.

As for prices, OpenAI models are enabling lower costs.

Data to Decisions Innovation & Product-led Growth Future of Work Tech Optimization Next-Generation Customer Experience Digital Safety, Privacy & Cybersecurity openai ML Machine Learning LLMs Agentic AI Generative AI Robotics AI Analytics Automation Quantum Computing Cloud Digital Transformation Disruptive Technology Enterprise IT Enterprise Acceleration Enterprise Software Next Gen Apps IoT Blockchain Leadership VR Chief Information Officer Chief Data Officer Chief Executive Officer Chief Technology Officer Chief AI Officer Chief Analytics Officer Chief Information Security Officer Chief Product Officer

Palo Alto Networks acquires Talon Cyber Security, Dig Security

Palo Alto Networks acquires Talon Cyber Security, Dig Security

Palo Alto Networks said it has acquired Talon Cyber Security, an enterprise browser security startup. The move comes days after the company acquired Dig Security.

Talon Cybersecurity aims to address attacks via unmanaged devices via its Talon Enterprise Browser. The Talon Enterprise Browser will be combined with Palo Alto Networks' Prisma SASE platform and look to protect unmanaged endpoints that connect to SaaS enterprise applications.

Dig Security is a startup focused on data security posture management, or DPSM.

Terms of the deal weren't disclosed, but TechCrunch put the figure at $400 million for Dig Security. Talon Cyber Security reportedly went for $625 million

The two deals highlight how Palo Alto Networks plans to acquire startups that can help build out its platform. 

According to Palo Alto Networks, generative AI adoption will require enterprises to take control of sensitive data across cloud services, databases, vector databases and platform as a service. Dig Security's technology gives enterprises the ability to discover, classify, monitor and protect sensitive data wherever it resides on the cloud.

Like Talon's technology, Palo Alto Networks said that Dig Security's DPSM platform will be integrated into its Prisma Cloud.

There's a race to create next-gen security platforms powered by AI between Palo Alto Networks, Crowdstrike, Zscaler and a host of others. 

Digital Safety, Privacy & Cybersecurity Security Zero Trust Chief Information Officer Chief Information Security Officer Chief Privacy Officer

Ignorance of AI is no excuse

Ignorance of AI is no excuse

Understanding and explaining the workings of artificial brains—particularly deep neural networks—has been a problem for a decade or so. Some AI entrepreneurs seem almost to boast they don’t know how their creations work, as if mysteriousness is proof of real intelligence. But algorithmic transparency is being mandated in new European legislation so that individuals have better recourse when adversely affected when robots miscalculate their credit or health insurance risks.

I want to discuss another reason regulators have for getting inside the black box of AI: accountability under data privacy regimes.

The power of conventional privacy laws

Large language models (LLMs) and generative AI are making it hard now to tell fact from fiction.  The Some commentators, with great care, call this an existential threat to social institutions and social order. Naturally there are calls for new regulations. Such reforms could take many years

But I see untapped power to regulate AI in the existing principles-based privacy laws that prevail worldwide, a famous example being Europe’s General Data Protection Rule (GDPR).

I have written elsewhere about the “superpower” of orthodox data privacy laws. These are based on the idea of personal data, broadly defined as essentially any information which may be associated with an identifiable natural person. Data privacy laws such as the GDPR (not to mention 162 national statutes) seek to restrain the collection, use and disclosure of personal data.

Generally speaking, these laws are technology neutral; they are blind to the manner in which personal data is collected.

This means that when algorithms produce data that is personally identifiable, those algorithms and their operators are in scope for privacy laws in most places around the world.

Surprise!

Time and time again, technologists are taken by surprise by the privacy obligations of automated personal data flows:

  • In 2011, German regulators found that Facebook’s photo tag suggestions violated privacy law. The company was ordered to cease facial recognition and delete its biometric data sets. Facebook prudently went further, suspending tag suggestions worldwide for many years. See also this previous analysis of tag suggestions as a form of personal data collection.  
  • The counter-intuitive Right to be Forgotten (RTBF) first emerged as such in the 2014 European Court of Justice case Google Spain v AEPD and Mario Costeja Gonzálezi.  Often misunderstood, the case was not about “forgetting” anything in general but specifically de-indexing web search results. The narrow scope serves to highlight that personal data generated by algorithms (for that’s what search results are) is covered by privacy law. In my view, search results are not simple replicas of objective facts found in the public domain; they are the outcomes of complex Big Data processes.

What’s next?

The legal reality is straightforward. If personal data comes, by any means, to be held in an information system, then the organisation in charge of that system may be deemed to have collected that personal data and thus is subject to applicable data privacy laws.

As we have seen, privacy commissioners have thrown the book at analytics and Big Data.

AI may be next.

Being responsible for personal data, no matter what

If a large language model acquires knowledge about identifiable people—whether by deep learning or the gossip of simulacra—then that knowledge is personal data and the model’s operators may be accountable for it under data privacy rules.

Neural networks represent knowledge in weird and wonderful ways, quite unlike regular file storage and computer memory. It is notoriously hard to pinpoint where these AIs store their data.

But here’s the thing: privacy law probably doesn’t care about that design detail, because the effect still amounts to collection of personal data.

If a computer running a deep learning algorithm has inferred or extracted or uncovered or interpolated fresh personal data about individuals, then its operator has legal obligations to describe the data collection in a privacy policy, justify the collection, limit the collection to a specific purpose, and limit reuse of the collected personal data. In the privacy laws I have read, there is nothing to indicate that an information system based on neutral networks will be treated any differently from one running written in COBOL and running on a mainframe. 

Privacy law usually gives individuals the right to request a copy of all personal data that a company holds about them.  In some jurisdictions, individuals have a qualified right to have personal data erased.

I am not a lawyer but I can’t see that owners of deep learning systems holding personal data can excuse themselves from technology-neutral privacy law just because they don’t know exactly how the data got there.  Nor can they logically get around the right to erasure by appealing to the sheer difficulty of selectively removing knowledge that is distributed throughout a neutral network. Such difficulty may be seen as the result of their own design and decision-making.

And if selective erasure of specific personal data is impossible with these black boxes, then the worst case scenario for the field of AI may be that data protection regulators rule the whole class of technology to be non-compliant with standard privacy principles.

Are you getting prepared for AI? 

Constelltion is developing new AI preparedness tools to help organisations evaluate the regulatory and safety implications of machine learning. Get in touch if you'd like to know more about this reserch, or to exchange views. 

New C-Suite Data to Decisions Digital Safety, Privacy & Cybersecurity Innovation & Product-led Growth Future of Work Tech Optimization Next-Generation Customer Experience ML Machine Learning LLMs Agentic AI Generative AI AI Analytics Automation business Marketing SaaS PaaS IaaS Digital Transformation Disruptive Technology Enterprise IT Enterprise Acceleration Enterprise Software Next Gen Apps IoT Blockchain CRM ERP finance Healthcare Customer Service Content Management Collaboration Security Zero Trust Chief Executive Officer Chief Information Officer Chief Technology Officer Chief AI Officer Chief Data Officer Chief Analytics Officer Chief Information Security Officer Chief Product Officer Chief Privacy Officer

Why generative AI workloads will be distributed locally

Why generative AI workloads will be distributed locally

This post first appeared in the Constellation Insight newsletter, which features bespoke content weekly.

Generative AI workloads have been dominated by Nvidia, a massive cloud buildout and compute that comes at a premium. I'm willing to bet that in a year, we'll be talking about distributed compute for model training and more workloads on edge devices ranging from servers to PCs to even smartphones.

On earnings conference calls, generative AI is still a common theme, but there's a subtle shift focusing on commoditization of the compute behind large language model (LLM) training and a hybrid approach that leverages devices that are being built with processors generative AI capable.

"There's a market shift towards local inferencing. It's a nod to both the necessity of data privacy and an answer to cloud-based inference cost," said Intel CEO Pat Gelsinger on the company's third quarter earnings conference call.

Here's a quick tour of what's bubbling up for local compute powered generative AI.

Amazon CEO Andy Jassy said:

"In these early days of generative AI, companies are still learning which models they want to use, which models they use for what purposes and which model sizes they should use to get the latency and cost characteristics they desire. In our opinion, the only certainty is that there will continue to be a high rate of change."

Indeed, the change coming for generative AI is going to revolve around local compute that's distributed.

Here's why I think we may get to distributed model training sooner than the industry currently thinks:

  • Enterprises are building out generative AI infrastructure that often revolves around Nvidia, who needs competition but right now has an open field and the margins to prove it.
  • The generative AI price tag is tolerated today because the low-hanging productivity gains are still being harvested. If you can improve software development productivity by 50% who is going to sweat the compute costs? Pick your use case and the returns are there at the moment.
  • But those easy returns are likely to disappear in the next 12 months. There will be more returns on investment, but compute costs will begin to matter.
  • Companies will also gravitate to smaller models designed for specific use cases. These models, by the way, will need less compute.
  • Good enough processors and accelerators will be used to train large language models (LLMs)--especially for cases where a fast turnaround isn't required. Expect AWS' Inferentia and Trainium to garner workloads as well as AMD GPUs. Intel, which is looking to cover the spectrum of AI use cases, can even benefit.
  • The good enough model training approach is likely to extend to leveraging edge devices for compute. For privacy and lower costs, smartphones, PCs and other edge devices are going to be equipped and ready to leverage local compute.

Ultimately, I wouldn't be surprised if we get to a peer-to-peer or Hadoop/MapReduce-ish approach to generative AI compute.

Data to Decisions Tech Optimization Innovation & Product-led Growth Future of Work Next-Generation Customer Experience Digital Safety, Privacy & Cybersecurity Big Data ML Machine Learning LLMs Agentic AI Generative AI AI Analytics Automation business Marketing SaaS PaaS IaaS Digital Transformation Disruptive Technology Enterprise IT Enterprise Acceleration Enterprise Software Next Gen Apps IoT Blockchain CRM ERP finance Healthcare Customer Service Content Management Collaboration Cloud CCaaS UCaaS Enterprise Service GenerativeAI Chief Information Officer Chief Technology Officer Chief Information Security Officer Chief Data Officer Chief Executive Officer Chief AI Officer Chief Analytics Officer Chief Product Officer

Disrupting yourself can hurt even if it makes sense, just ask Paycom

Disrupting yourself can hurt even if it makes sense, just ask Paycom

A common mantra among technology companies is that it's better to disrupt yourself than let a competitor do it. On the whiteboard that mantra makes long-term sense. In reality, a business model transition can crush your stock.

Just ask Paycom, a human capital management software provider that created Beti, a service that dramatically reduces payroll errors and drives value for customers. The concept is shockingly simple: Give employees access to payroll to fix errors before the checks are cut.

Paycom shares were hammered this week because Beti is doing away with unscheduled payroll runs and error fixes. What's wrong with generating returns for customers? A more perfect payroll means fewer billable items for Paycom, which charged for corrections and unscheduled payrolls.

Here’s a look at the carnage (hint: lower right on chart is where “PAYC” wound up.)

To its credit, Paycom is playing the long game with customers. Sure, the business model transition hurts a bit, but other companies have gone through it. Think about how software companies transitioned from licenses to subscriptions after Adobe led the way.

Beti launched in 2021 and now two-thirds of Paycom's customer base is using it. Paycom is rolling out Beti to Mexico to complement US and Canada.

Paycom CEO Chad Richison explained the Beti effect.

"For most employees, the value of the perfect payroll is oftentimes immeasurable. If their check is perfect, they don't need to borrow money from a friend or family member to get through the weekend or make a bill payment. How do you measure the value of that?

We're getting better and better at helping employers measure the full value available to them when payrolls are perfect. A portion of that value is easy to calculate because it's the value they receive by the elimination of after-the-fact payroll errors that require correction payroll runs, manual checks, voided checks, direct deposit reversals, additional wires, tax adjustments, W2Cs, et cetera, et cetera.

Perfect payrolls eliminate these common after-the-fact payroll corrections that would otherwise be billable. So the more employees do their own payroll, the greater the savings delivered to the client from Paycom future billings, which results in lower related revenue recognized by Paycom."

Paycom isn't hurting. The company delivered third-quarter revenue of $406 million, up 22% from a year ago, with net income of $75 million, or $1.30 a share. Non-GAAP earnings were $1.77 a share. Third-quarter sales missed guidance because Beti was working too well. In addition, the outlook was light as Paycom projected revenue between $420 million to $425 million in the fourth quarter.

In other words, the more customers use Beti, the more value they receive at the expense of Paycom revenue. Paycom is expecting revenue growth in 2024 between 10% and 12%. A customer that had to ran 19 payrolls a quarter due to errors can now run 13.

The big question: Should Paycom have disrupted its own services business? It's a conundrum faced by many companies and the decision isn't easy. Here are a few thoughts about answering that question.

Paycom's long game. Richison said that he's focused on "the client value and the differential between what they're paying and what they're actually achieving." Overall, I don't think you can go wrong providing value to customers.

If Paycom didn't launch Beti someone else would have. I'd argue that payroll errors aren't a feature but a bug. Some startup would have disrupted the error-ridden payroll process with something similar to Beti anyway. Paycom would have lost the services revenue and customers too over time. Shortly after Beti launched, Richison said on Paycom's fourth quarter 2021 conference call:

"For years, I have been predicting the end of the old model whereby HR and payroll personnel’s routine of inputting data for employees is replaced by a self-service model that provides employees direct access to the database.

The old model is dying and that is good for both the business and the employee."

Here's a fun fact: Paycom's fourth quarter revenue in 2021 was $285 million, well below the projected $420 million or so two years later.

Don't forget word of mouth. Paycom received a good amount of attention as shares fell this week. Enterprise buyers who dig a bit will quickly find a "man bites dog" headline as Paycom sacrificed revenue for customer value. Something tells me this Paycom stock tale may wind up being good marketing.

Paycom is adding new customers. Should Beti ramp even further Paycom will have more enterprise customers on its platform. These customers tend to buy other services later if the vendor delivers.

"New business sales as well as cross-selling within our base has always been a mitigating factor to any type of transition shift, we make like this," said Richison. "New business sales remained strong. In fact, most of the calls we get in are about Beti. We've got our first enterprise rep and they're only targeting deals that have greater than 25,000 employees. And they've got plenty of leads."

Could Paycom have managed this transition better? Perhaps. But Beti appears to be a hit. In other words, take the win with a bit of pain. Disrupt yourself or be disrupted.

Future of Work Data to Decisions Innovation & Product-led Growth Tech Optimization Next-Generation Customer Experience Digital Safety, Privacy & Cybersecurity ML Machine Learning LLMs Agentic AI Generative AI AI Analytics Automation business Marketing SaaS PaaS IaaS Digital Transformation Disruptive Technology Enterprise IT Enterprise Acceleration Enterprise Software Next Gen Apps IoT Blockchain CRM ERP finance Healthcare Customer Service Content Management Collaboration Chief Executive Officer Chief Information Officer Chief Technology Officer Chief AI Officer Chief Data Officer Chief Analytics Officer Chief Information Security Officer Chief Product Officer

Google Cloud CEO Thomas Kurian on DisrupTV: Generative AI will revamp businesses, industries

Google Cloud CEO Thomas Kurian on DisrupTV: Generative AI will revamp businesses, industries

Google Cloud CEO Thomas Kurian said companies are starting to reimagine their businesses for artificial intelligence one process at a time. The big question is what industries hit scale first. 

Speaking on DisrupTV, Kurian said "in virtually every field, we're seeing people take the persona that they had, and then creating a digital version of that persona using AI and this is happening incredibly quickly."

"The common theme is taking the skills of the model that represents what human beings can do, but now creating a digital persona and assisting a function," he said.

Kurian cited use cases across industries including healthcare, insurance, cosmetics and media to name a few. Many of these industries started out with AI use cases that revolved around classification and categorization. The next phase for AI was prediction using models based on different parameters. Generation is the new phase.

"Generation is the next skill as you train a model with a set of inputs and it can generate output," said Kurian. "You can now put together and automate a complete workflow for the whole company. And this is happening in many places. We see it happening at scale at many around the world."

Related: Cloud customers still optimizing spend and should forever | Google Cloud delivered third quarter revenue of $8.4 billion, up 22% from a year ago, with an operating profit of $266 million

Constellation Research CEO Ray Wang asked Kurian whether incumbent companies or startups had advantages in generative AI. Kurian said:

"We're very early in the market. To do AI well, we need high quality data sets to fine tune the state-of-the-art models. For high quality datasets, you need state of the art models, and you obviously need the infrastructure to serve the models. But then just as importantly, you need to integrate these into the application surface.

The companies that succeed will have capabilities for state-of-the-art models that are driven by high quality datasets that they own, and the ability to activate these models. within account within the context of an application surface.

There will obviously be disruptors and they will take a function of process that was done in one way and fundamentally change it. You will see disruptions in different industries, where the fundamental business model itself may change.

"The winners are always those that solve a fundamental problem materially."

Going forward, generative AI will be democratized and simplified. AI will be an enabler for new technologies and access. In other words, we'll all become programmers to some degree. Kurian said:

"The code generation model will generate the skeleton code and create the environment. You can say 'hey, I want to support a million and a half users with less than 10 seconds latency of 10 milliseconds latency and I need to guarantee four nines of availability' and the system will do it for you. There's no reason that that cannot be done with where models are. And by doing that, we again change how widely accessible these things are. We are also very encouraged by the fact that we can make these models work not just for people in affluent countries, but also people in emerging markets."

Kurian also touched on other topics.

Security. Kurian said AI is already detecting threats, but also prioritizing them. "We built a model that can look at the threats that are emerging in your infrastructure, which one needs priority and how does that threat affect you," he said. "Models don't have that emotional bias. So, they can look at many more patterns to detect what's going on what is the attack the attack surface. It can remediate it and automate the creation of the rulebook to resolve the problem. We're applying AI to the whole spectrum."

AI-driven cybersecurity attacks. "We also see AI being used by bad actors to create new types of threats and we're also building our platform to thwart new types of threats," he said.

Running Google Cloud. Kurian said Google had great products but needed to build out enterprise capabilities as an independent unit.

"We're the fifth largest software company in the world, which is a long, you know, huge credit to the team. But when we looked at it, we felt we needed to do four things really well. You need to take great technology but convert it into solutions that people can use. Just having technology that's not accessible is a challenge.

Second, we need to build a great go to market function. What kind of structure do you have? How do you focus? We started with a set number of industries and countries.

It's an ecosystem game. It's not your company against another company. It's your ecosystem. So, we made decisions very early to partners. We started with 100 partners today there are 100,000 partners. And part of that is we wanted to bring that ecosystem so that people realize it's a bigger pie that they are creating, not slicing off the same pie. That's the third one.

And then you have to do something really well. It's just like sports. In order to play really well, you have to do the grunt work of training. We have to do a lot of the things below the surface of the water such as the systems, the legal contracting, and the frameworks to be more efficient as an organization. Those were all put in place so that you can go faster. Unless you have a strong core, you can't really play well. We've been super fortunate that we've been blessed with such a great team of people that have done so much of the work to get us where we are today."

Data to Decisions Tech Optimization Digital Safety, Privacy & Cybersecurity Innovation & Product-led Growth Future of Work New C-Suite Next-Generation Customer Experience Google Google Cloud SaaS PaaS IaaS Cloud Digital Transformation Disruptive Technology Enterprise IT Enterprise Acceleration Enterprise Software Next Gen Apps IoT Blockchain CRM ERP CCaaS UCaaS Collaboration Enterprise Service AI GenerativeAI ML Machine Learning LLMs Agentic AI Analytics Automation Chief Information Officer Chief Technology Officer Chief Information Security Officer Chief Data Officer Chief Executive Officer Chief AI Officer Chief Analytics Officer Chief Product Officer

Apple Q4 better than expected, but it has a Mac sales sink

Apple Q4 better than expected, but it has a Mac sales sink

Apple's fourth quarter results were better-than-expected but revenue was down for the fourth consecutive quarter. Mac sales in the quarter were weaker than expected, but may get a lift with new MacBook Pro models on tap. 

The company reported fourth quarter earnings of $1.46 a share on revenue of $89.5 billion, down 1% from a year ago.

Wall Street was expecting Apple to report fourth quarter earnings of $1.39 a share on revenue of $89.28 billion.

Here's a breakdown of Apple fourth quarter results by product line and their targets via LSEG.

  • iPhone revenue: $43.8 billion; Estimate: $43.81 billion
  • Mac revenue: $7.61 billion; Estimate: $8.63 billion
  • iPad revenue: $6.44 billion; Estimate: $6.07 billion
  • Wearables, Home and Accessories revenue: $9.33 billion; Estimate: $9.43 billion
  • Services revenue: $22.31 billion; Estimate: $21.35 billion

In a statement, CEO Tim Cook said the company has a strong lineup for the holiday season including new Macs. Overall, Apple's game plan is to monetize its user base with services and subscriptions.

For fiscal 2023, Apple reported net income of $96.99 billion, or $6.13 a share, on revenue of $298.08 billion.

Other items to know:

  • China revenue in the fourth quarter was $15.08 billion, down from $15.47 billion a year ago.
  • Americas revenue was $40.11 billion, up from $39.8 billion a year ago.
  • Europe and Japan revenue was down slightly from a year ago.

Constellation Resarch CEO Ray Wang said:

"Despite continued revenue decline, Apple is a digital giant and flight to safety stock in good times and bad. The macro conditions are elongating the iPhone replacement cycle. China is the challenge as iPhone 15 sales slow and Huawei has revamped its offerings. The elongation of iPhone replacement cycles is the headwind. 

The new Mac Lineup provides cost savings and higher margins and could revitalize sales. The vertically integrated strategy is working For Mac, iPhone, Watch, and ultimately Vision Pro. Services is the bright spot. All eyes on the holiday forecast for the December quarter." 
 

More:

Next-Generation Customer Experience apple Chief Information Officer

Palantir's commercial business scales with help of AI boot camps

Palantir's commercial business scales with help of AI boot camps

Palantir's commercial annual revenue run rate is closing in on the company's government business as Palantir Artificial Intelligence Platform (AIP) gains enterprise traction.

The upcoming parity point between commercial revenue and Palantir's core government business is worth watching. In the third quarter, Palantir commercial revenue grew 23% to $251 million with US revenue growing 33% to $116 million. Government revenue grew 12% to $308 million in the third quarter.

Overall, Palantir reported third quarter earnings of $72 million, or 3 cents a share, on revenue of $558 billion, up 17% from a year ago. Adjusted earnings were 7 cents a share. The results handily topped expectations.

Palantir has been trying to grow its enterprise revenue base for years and now has 181 commercial customers, up 37% from a year ago. As for the outlook, Palantir is projecting 2023 revenue between $2.216 billion and $2.22 billion. J.D. Power, Palantir team up on generative AI apps for auto value chain

On a conference call with analysts, Ryan Taylor, Palantir's Chief Revenue Officer, said the company closed 80 deals including 12 worth $10 million or more across 11 industries.

Taylor said:

"We're also seeing the acceleration of larger deals and shorter times to conversion and expansion, including a multiyear deal in excess of $40 million with one of the largest home construction companies in the U.S. to start up pilot and converted all within Q3. This growth is in part due to AIP's continued transformation of the way we partner with and deliver value for our customers, and we expect AIP's impact to continue to intensify."

Palantir has seen traction with "AIP boot camps," which deliver real workflows on customer data in 5 days or less. That approach is driving contract expansions. "We're on track to conduct boot camps for more than 140 organizations by the end of November, nearly half of those are taking place this month alone, which is more than the number of U.S. commercial pilots we conducted all of last year," said Taylor. "We almost tripled the number of AIP users last quarter and nearly 300 distinct organizations have used AIP since our launch just five months ago. We will continue investing meaningfully in boot camps as our go-to-market strategy for AIP."

Of course, Taylor noted that Palantir's government business remains strong and should accelerate going forward.

Shyam Sankar, Palantir's CTO, said AIP boot camps are driving the point home that you can't use LLMs without tools to provide algorithmic reasoning. Sankar recently gave a talk on the topic.

According to Palantir CEO Alex Karp the commercial success isn't surprising if you zoom out and consider the company's military experience. Karp said:

"AIP and U.S. commercial is not only is disrupting the market, it's setting a standard that I don't believe any other software company will be able to reach partly because they misunderstood the value of LLMs and their relative importance, and lack of importance, partly because they don't have decades of experience on the frontline as we do in the military with managing the core ways in which you make these things precise, the way in which you provide governance."

Karp added that AIP enables enterprises to manage LLMs and "basically pen test your enterprise." "My view of what we should do is build products that are so good that the competition stops competing, whether that's in commercial or on the battlefield and that's what we're doing. And that's what we're seeing in AIP," said Karp.

Data to Decisions Innovation & Product-led Growth Future of Work Tech Optimization Next-Generation Customer Experience Digital Safety, Privacy & Cybersecurity ML Machine Learning LLMs Agentic AI Generative AI AI Analytics Automation business Marketing SaaS PaaS IaaS Digital Transformation Disruptive Technology Enterprise IT Enterprise Acceleration Enterprise Software Next Gen Apps IoT Blockchain CRM ERP finance Healthcare Customer Service Content Management Collaboration Chief Information Officer Chief Data Officer Chief Executive Officer Chief Technology Officer Chief AI Officer Chief Analytics Officer Chief Information Security Officer Chief Product Officer

SAP Build Code adds generative AI tools, bridges to process data, Java, JavaScript

SAP Build Code adds generative AI tools, bridges to process data, Java, JavaScript

SAP launched SAP Build Code development tools that leverage generative AI and court Java and JavaScript developers as well as connect to data stores across the enterprise.

At SAP's TechEd 2023 conference, SAP outlined a series of tools that aim to enable developers to build applications and connect them to the SAP Business Technology Platform. These pro code tools, which build on last year's low code/no code rollout, feature the following:

  • Generative AI productivity tools to connect heterogenous systems.
  • Vector embedding capability in SAP HANA Cloud to handle unstructured data and provide context to models.
  • SAP AI Foundation, a developer repository to build AI-driven business apps.
  • Development tools that are optimized for Java and JavaScript and interoperate with ABAP Cloud development.
  • Code generation via SAP's generative AI assistant Joule.
  • Unification of design and run-time services to build interfaces using Java and Node.js.
  • The ability to work in Visual Studio cloud hosted environments.
  • Simplified API management so developers can connect to third party systems, link processes and keep business context.

SAP Build Code will be generally available in the first quarter.

The ERP giant's other big play was to connect its SAP Build Code developer base to its process optimization tools including SAP Build Process Automation, SAP Signavio, which will get a large language model specific to business processes or large process model, and SAP Integration Suite. By connecting development and process optimization, SAP is looking to keep data on its platform while extending into other third-party systems. To that end, SAP also said that models built in DataRobot can be hosted inside SAP AI Core and used in SAP Build applications and extensions.

More:

In addition, SAP Datasphere will get updates to import business context from SAP S/4HANA Cloud, SAP S/4HANA, SAP Business Warehouse and SAP BW/4HANA for data integration. Datasphere will also have simplified data sharing as well as prebuilt business content with a release planned for the fourth quarter.

Constellation Research's take

Holger Mueller, Constellation Research analyst, handicapped SAP's announcements. He said:

"SAP is doing right by its developer community with the launch of Build Code – which took a temporary back seat with all the attention the low code/no code offering had in the last 12 months. It is of strategic importance for both SAP customers and SAP alike that existing customers can move their ECC based customizations and extensions to the cloud future, where they need to upgrade to S/4HANA by 2027.

Given that most custom SAP code is in ABAP though, the SAP developer community wonders why the focus was Java/JavaScript. Certainly, a more en vogue platform today, but it is easier to move alike code / programming language – than re-writing. It would have been a big win if SAP could have delivered all pro-code tooling in 2023.

On the HANA side it is important that HANA gets the now 'standard' AI vector feature – what it will mean for in memory usage will have to be seen – but a key step to keep HANA and HANA data (where all of S/4 HANA data is) relevant in the AI age. What needs to be clarified is the basic working of SAP’s AI platform – that would have been a prime-time topic for TechEd. We may now have to wait till Sapphire."

Data to Decisions Tech Optimization Digital Safety, Privacy & Cybersecurity Innovation & Product-led Growth Future of Work Next-Generation Customer Experience SAP AI GenerativeAI ML Machine Learning LLMs Agentic AI Analytics Automation Disruptive Technology Chief Information Officer Chief Technology Officer Chief Information Security Officer Chief Data Officer Chief Executive Officer Chief AI Officer Chief Analytics Officer Chief Product Officer