- When Vishal Sikka joined the Oracle board of directors, Oracle claimed he was one of the world’s leading experts in AI.
- We analyze whether this claim by Oracle is true.
When Vishal Sikka joined Oracle’s board of directors, some amazing things were claimed and repeated throughout the media.
Notice this web page from the Economic Times of India. He is called one of the world’s leading experts in artificial intelligence.
However, who actually made this claim? Well, it turned out to be Larry Ellison.
“Vishal is one the world’s leading experts in Artificial Intelligence and Machine Learning,” said Oracle Chairman and CTO Larry Ellison. “These AI technologies are key foundational elements of the Oracle Cloud’s Autonomous Infrastructure and Intelligent Applications. Vishal’s expertise and experience makes him ideally suited to provide strategic vision and expert advice to our company and to our customers. He is a most welcome addition to the Oracle Board.” – TechR
First things first.
- Oracle has no known AI capabilities and is not known for AI. Oracle, just like other vendors, uses AI as a marketing construct.
- As we covered in the article How Real is Oracle’s Autonomous Database?, the autonomous database is not real and is simply a way of keeping pace with AWS’s RDS database service.
- Oracle has very little cloud business and is also not considered competent in the cloud, with engineers working in the background to provide faux cloud.
- Oracle also does not have “intelligent applications.” Oracle has several applications that they acquired because Oracle is not capable of developing effective applications internally.
These are critical points because Larry Ellison tells multiple lies in just one paragraph. So before we get to Larry Ellison’s claim around Vishal Sikka, we should first appreciate that everything that Larry Ellison has said in this paragraph is false. Obviously, this does not bode well for anything else we will fact check. But who knows, maybe Larry inserted one true thing in the paragraph.
Larry’s Unending Verbal Diarrhea Of Lies
Now let us get to why Larry Ellison has told so many lies in this paragraph.
Well, Oracle is using trying to get maximum PR benefit from the announcement, and they know this announcement will be carried and repeated uncritically in many outlets. Therefore, the strategy is to incorporate as many marketing elements into the PR release as possible. Most media outlets employ journalists who have never worked in IT, and they are will repeat whatever is in the press release.
Hence we get to the claim about Vishal Sikka supposedly being a world expert in AI. Notice that The Economic Times of India did not say who said this. It simply included the statement so that it seems as if it is a claim made by India’s Economic Times.
Oracle marketing must have been over the moon when they saw how uncritically this claim was accepted!
When you get inept and deadline-constrained media entities who think their job is done when they obtain a quote to repeat your marketing messaging for free, that is referred to as a Crystal Moment.
Furthermore, virtually the same story with the same quotes was carried in multiple media outlets, with no media entity analysis. Because so much of the article was simply quotes from Oracle, these articles can be considered simply written by Oracle. This issue of mindless media entities repeating the same press releases was something we specifically noted in the article How Awful Was the Coverage of the McDonald’s AI Acquisition?
No Effort, No Thinking
These “no effort and no thinking” articles help media entities get clicks with minimal investment. Still, it also makes it look like there is consensus on a topic, when in fact, a large number of media entities have copied and pasted the same press release onto their website. IT media entities know that entities like Oracle have a sophisticated marketing department adept at manipulating them. It is not responsible for serving as their passive repeaters, as it takes advantage of the reader’s naivete and distributes false information.
Evidence of Vishal Sikka’s Expertise in AI
Vishal Sikka began making claims around being an AI expert when he was raising money for his new AI venture, Vianai, where he raised $50 million. However, I have been following Vishal Sikka for several years as I analyze and fact-check SAP and Vishal Sikka was previously the CTO of SAP.
What struck me as odd is that I do not ever recall once hearing about Vishal Sikka within the context of AI until very recently. Vishal Sikka’s dissertation is in AI. However, Vishal Sikka is 52 years old. And his work experience shows virtually no AI since that time.
This is not unusual.
Before AI became hot again in the past 4 to 5 years, it was widespread for there to be few jobs in AI when Vishal graduated back in 1996, and so people that had degrees in AI moved over into other areas of computing. There are also many people with degrees in robotics who switched into general systems implementation or development.
It just so happens that AI is now hot.
Where Did All This AI Work Experience Come From?
Normally to be considered a world expert in something, one needs to actually work in the field, normally for a few decades. However, Vishal Sikka never did this. Vishal Sikka’s work experience shows a person who is both a technology generalist and who has been in very senior roles for over 10 years, including leading a 225,000 person company in Infosys — which is also not known for its AI work. A second point. I spoke to several people I consider AI experts but who are not in the SAP space. None of them had heard of Vishal Sikka.
I performed a search in an academic search engine, and Vishal has no published articles on AI (since his 1996 dissertation). All of Vishal’s publications are associated with SAP Labs, which does not publish research as much as marketing material in a research format — but they have multiple authors, and it’s unlikely much of it was Vishal’s work.
An Amazon book search shows no books on AI and no books of any kind. Nothing.
Vishal Sikka is a world expert in a topic without a single publication in the topic after his dissertation? This has to be the strangest world expert I have ever heard of. Perhaps he is a secret world expert?
In reviewing his Wikipedia entry, Vishal did not work on AI from the time he began at SAP in 2002, and nor would he have had much to do with AI when he was the head of Infosys for 3.5 years.
Vishal Sikka was who Hasso Plattner tapped to complete a flawed database design, based upon Hasso Plattner’s massive overestimation of his own capabilities to design a database. For several years, Vishal Sikka was presented to audiences as a deep expert in databases. But it is also not clear where Vishal Sikka acquired this expertise, as he had not worked in databases before being tasked with recovering the HANA project. This video shows a person who uses the term “fundamentally” frequently but does not appear to have an excellent grasp of what he is talking about. SAP has added precisely zero to the intellectual property of databases since HANA was introduced. SAP has been reverse engineering other far more advanced databases, as we cover in the article Did SAP Reinvent the Wheel with HANA?
At SAP, “enormous innovation” amounts to backward engineering other company’s products. This is something I noted back in 2010 in the article It’s Time for the xApp Program to End, and long before Teradata filed its lawsuit against SAP for both anticompetitive behavior and theft of its data warehouse IP (where Vishal Sikka is named as a complicit party).
Specific Topics from the Video
- Aggregates are a Problem?: Vishal Sikka’s statements about not requiring pre-aggregates (what is a pre-aggregate Vishal?) and closed systems are really just babbling designed to sound good but contain little actual content. Pre-aggregates or just aggregates are a good part of database design, and many aggregates don’t change and don’t need to be recalculated.
- HANA is Moving Away from Closed Systems?: SAP makes the most closed systems in enterprise software, even having the most punitive controls over accessing their systems, called indirect access. So there is no more closed set of systems than those offered by SAP. HANA has its own indirect access liabilities and often requires a second HANA instance as HANA cannot be connected to other systems without incurring indirect access liabilities, as we cover in the article The HANA Police.
- Simplification?: HANA is not a simplification of anything. HANA does not have a more simple data model, as we cover in the article Does S/4HANA Have a Simplified Data Model? HANA is the highest maintenance overhead database that we track (which can be seen at The Brightwork Comparison and Scoring of Databases) and marred by continual maturity.
- HANA Replacing Data Marts and Data Warehouses?: SAP customers did not replace data warehouses and data marts with HANA. HANA is a database, not a data warehouse or a data mart. SAP’s go-to data warehouse/data mart solution is the BW, and companies ported their BW instances to HANA. Nothing of what Vishal is discussing in this area of the video is accurate.
- Moving to a Single HANA Instance?: One of Vishal’s more hallucinatory proposals was that all applications and analytics would run from a single HANA instance. Did we cover this fallacy in the article SAP’s Big Idea: ERP Data and Big Data on the Same Database?
- Massive Performance Improvement?: Vishal claims that a process that previously took 3 days in ERP on Oracle was reduced to 2 seconds on HANA. Let us do the math on this. 3 days × 24 hours per day× 60 minutes per hour × 60 seconds per minute = 259,200 seconds. 259,200 / 2 = 129,600 times faster. This calculation was for an incentive program for loyal customers, not even in HANA’s wheelhouse of reading performance. It is simply impossible for this claim to be true unless there was something else changed, some misconfiguration that Vishal is not including.
This quote encapsulates his view and is a prediction.
I believe in a completely different reality which is significantly more simplifying than what we have used to is in front of us. That is the opportunity that we are after with HANA. I think that reassembly of existing components and layers towards that misses the point. – Vishal Sikka
Eight years after this video was made, there is very little difference in SAP projects with HANA than when this video was filmed, and these grand projections were made. Analytics run faster, but not any faster than any of the competing databases. SAP’s claims about everything “in memory” never came true because HANA is just another memory-optimized database, although with an overly expensive hardware specification.
Hasso Made Him Do It?
The only “out” that Vishal has here is that he has to repeat whatever Hasso Plattner says, and Hasso Plattner defined the language around HANA before Vishal inherited HANA. With Hasso Plattner, you either do exactly as he says and pretend he is a genius, or you find yourself another job.
Furthermore, none of the things proposed by SAP about HANA’s superiority regarding competing databases turned out to be true. We analyzed claims made by SAP and John Appleby of Bluefin Solutions in the article The Appleby Accuracy Checker: A Study into John Appleby’s Accuracy on HANA.
Therefore, Vishal Sikka was previously presented to the general public as a database expert when he was not one. And he did nothing to make HANA a more effective database, yet he constantly receives credit for HANA by journalists, who don’t have the background or interest in fact-checking what happened with HANA. Brightwork Research & Analysis has performed the most independent research on HANA. Unlike virtually all media entities in IT, we take no income from SAP or any other vendor, and we have concluded that the claims around HANA by Vishal Sikka and SAP are false. And that they must have known they were false at the time they made them.
A Lack of Fact-Checking by Media Entities
In fact, notice this statement from Quartz India.
He is known as the “father of HANA,” SAP’s fastest-growing product ever—in 2014, this platform was generating over $1 billion in revenue.
Being the HANA father does not even fit with SAP’s official story of HANA, which we cover in the article, Did Hasso Plattner and His Ph.D. Students Invent HANA?
Secondly, this writer simply repeats false statements made by SAP about HANA, which we fact-checked in the article How Accurate Was SAP on HANA Being the Fastest Growing Product In SAP’s History?
The question on a lot of people’s minds these days is whether SAP HANA will be as big as R/3? Plattner has said several times he believes it will be—while SAP will continue to support its current database partners, it will optimize all of its applications to run on the HANA database. And at TechEd this year, Sikka and SAP offered more details. – ASUG
This is why it is a good idea not to read ASUG (unless you perform media criticism as I do). You end up getting highly slanted information. ASUG is a captive entity to SAP and presents itself as a user group. At the same time, it primarily serves as a marketing and customer control arm of SAP, as we cover in the article How ASUG Lost its Way and Sold Out to SAP.
HANA was released in 2011. That has given it over 9 years to grow. If HANA has been the fastest-growing product in SAP’s history, then why is it ranked next to FileMaker according to DB-Engines?
Even a person who works in SAP and has no special research capability knows quite well that the fastest growing product in SAP’s history was R/3.
Vishal’s Legacy of Inaccuracy
I have spent a great deal of time reading quotes from Vishal Sikka, and he tends to speak at a very high level when discussing technology topics. I don’t think of technology insights when I think of Vishal Sikka. When I analyze Vishal Sikka’s statements, the most prominent thing that comes to mind is that the things Vishal Sikka says don’t come true.
Here is Vishal Sikka proposing to Denis Howlett (of Diginomica and a recipient of SAP funding) that SAP was putting a lot of emphasis on addressing cancer by running genetic algorithms in HANA.
What is curious is that nothing stated here came true. SAP’s supposed health care solution never became anything used by SAP customers. We covered Vishal Sikka’s inaccuracy in the article How Accurate Was Vishal Sikka on HANA’s Future?
So far, we are the only entity to go back, and fact checks whether anything stated here came true. There is also little validity to Vishal Sikka’s statements about running genetic algorithms in the database, as there is plenty of time for processing in medical research. These computers are dedicated to these tasks. The issue is the time required by humans to analyze the outcome.
Vishal Sikka is not occasionally wrong. Every time we have checked what happened versus what Vishal Sikka said would happen, Vishal turned out to be wrong. We covered his prediction about Sybase in the article How Accurate Was SAP on the Sybase Acquisition?
Review the accuracy of Vishal Sikka on the Sybase acquisition for yourself.
Safra Catz Adds On to Larry Ellison’s False Claims
As I brought up previously, Oracle wants to wring the most PR value out of its announcement of Vishal being appointed to the board.
“The digital transformation of an enterprise is enabled by the rapid adoption of modern cloud applications and technologies,” said Oracle CEO Safra Catz. “Vishal clearly understands how Oracle’s Gen2 Cloud Infrastructure, Autonomous Database and Applications come together in the Oracle Cloud to help our customers drive business value and adapt to change. I am very happy that he will be joining the Oracle Board.” – TechR
Here we have the same strategy, say that the person you are appointing knows basically “how great you are.”
Vishal Adds On His False Claims
“For years, the Oracle Database has been the heartbeat and life-blood of every large and significant organization in the world,” said Dr. Vishal Sikka. “Today, Oracle is the only one of the big four cloud companies that offers both Enterprise Application Suites and Secure Infrastructure technologies in a single unified cloud. It’s unique position in both applications and infrastructure paves the way for enormous innovation and growth in the times ahead. I am excited to have the opportunity to join the Board, and be part of this journey.” – TechR
Here Vishal Sikka does some marketing for Oracle, and like the other people in this press release, he is providing false information. First, as brought up by Markian Jaworksy..
“The big four cloud companies” 🙂 Everyone knows there are only 2 that are big.
And this is quite true. As Oracle has a negligible cloud business, how are they one of the four big cloud companies? So the statement by Vishal is false. As Oracle is not one of the four big cloud companies, it is not..
only one of the big four cloud companies that offers both Enterprise Application Suites and Secure Infrastructure technologies in a single unified cloud
Vishal has only just begun making public statements as a member of Oracle, and he is already providing not slightly but highly false information.
Vishal also thinks that Oracle is in a unique position in both applications in infrastructure.
Did Vishal Sikka think this back when he worked as the CTO for SAP?
I ask because I can’t find Vishal Sikka’s instances praising Oracle back when he worked for SAP.
Has Vishal told Larry what he thinks about Exalytics?
Because in the video included above, Vishal stated in the video the following.
Based upon what I read at least, I don’t think that Exalytics is worth criticizing. It misses the whole point of the new real time that we believe new enterprises are looking for.
Apparently, Vishal does not think companies should be buying either Exalytics or Exadata.
Vishal’s View on HANA Versus Oracle?
Vishal also did not think much of Oracle 12’s in-memory capability, stating in September 2013 that..
“Well, they are right that it is not comparable to HANA,” Sikka said in an emailed statement.
“The in-memory option they announced completely misses the point because the column store is read-only, and therefore it is a redundant replica of data in the row store, which means redundant storage and at least 5x larger size.”
“Plus, it hasn’t actually been released yet, so please come back and talk to us when it is available,” Sikka added. – ComputerWorld
Furthermore, Vishal stated about Exadata..
SAP HANA offers more simplification, while Exadata offers more layers. “As businesses attack more-complex problems, they need to rethink all the layers involved in their technology stacks,” said SAP’s Amit Sinha, vice president of Solution Marketing for SAP In-Memory Computing, in a recent conversation. “From the beginning with HANA, Vishal [Sikka] has talked about the need for dramatic simplification. But with Exadata, you still get more layers—it may be faster than your old database structure, but you’ve also got more backup layers, more replication layers, and more of the old stuff that just adds cost and complexity. And cost-shifting is definitely not the same as cost-optimization.” – Forbes
Does Vishal still have such a diminished view of Oracle’s database and data appliance offerings? Becuase it was extremely negative on them.
How about HANA? Does Vishal still think that most HANA is so advantageous that SAP accounts should migrate away from Oracle and HANA? Or has this view changed along with his income and position with Oracle?
If it has changed and thinks SAP customers should stay with Oracle and not move to “his baby” HANA, he was lying back when he told everyone that the Oracle database needed to be replaced?
SAP Admitting They Took IP from MySQL and Not Teradata
In the court case of Teradata v. SAP, SAP stated in court documents that they took IP from MySQL and not Teradata, as claimed by Teradata. We cover this in the article How SAP Admitted in Court Documents to Copying from Open Source. This filing would seem to greatly interest Oracle, as they now own MySQL and are fresh off of suing Google over Java.
Vishal Sikka is not a world expert on AI. In fact, he is not even prominent in the space.
Vishal Sikka has a Ph.D. in AI but has not done even close to enough work to be considered anywhere close to what Larry Ellison claimed. This claim by Larry Ellison is yet another attempt to associate Oracle with AI. Who knows if Larry Ellison even believes it. But what Larry Ellison got many media entities to publish is that Oracle has nabbed one of the world’s experts in AI.
We give Larry a 1 out of 10 for accuracy in this claim and award him our prestigious Golden Pinocchio Award.
Curiously, we read every article that provided coverage of Vishal Sikka’s addition to the Oracle board, and we did not find a single article that questioned whether Vishal Sikka was a world expert on AI. Does this mean that the IT media system is credulous and unreliable? Yes, yes, it does mean that. Because it is not just this one case, but it is now very standard for IT media to republish press releases without analyzing what they are publishing.
The Real Reason Vishal Sikka is Back Talking About AI
The only reason that Vishal Sikka has suddenly positioned himself as an AI expert is that AI is known as the easiest area to raise venture capital. We consider Vishal Sikka to be a highly unreliable source of information. Larry Ellison has been known to throw enormous sums of money at people he thinks have AI experience, as we mistakenly believe AI is critical to Oracle’s future. At one point, Oracle was paying Vishal Sikka $2000 per hour for his “AI expertise.”
However, in the paperwork announcing Sikka’s appointment to its board, Oracle did disclose that it signed a three-month contract with Sikka’s consulting company, Hang Ten, that pays $2,000/hour. Oracle says the contract will involve between 20 – 50 hours of consulting services and sales support per month, which works out to $40,000 – $100,000 a month. That’s on top of the pay and stock Sikka will receive as an Oracle board member. – TechR
This is surreal, and it illustrates that people with very marginal expertise can raise $50 million in venture capital or $2,000 per hour if they have the right branding. Furthermore, the accuracy of previous statements is, for all intents and purposes, meaningless.