- Vishal Sikka makes a number of false statements about HANA in this interview with Dennis Howlett.
- One of these is a rather hilarious proposal around curing cancer with genetic algorithms in HANA.
The video below was posted in 2012 and is an interview with Denis Howlett and Vishal Sikka.
In this article, we will review how accurate this video was, now that 4.5 years have passed.
Vishal may have been a Ph.D., but everything he said in this and other quotations from him has turned out to be false. Dennis Howlett, whose media entity receives income from SAP and who has a long pattern of not contradicting what SAP says, shows a clear deference to Vishal Sikka, even though even when this interview was performed, Vishal already had a long history of providing false information.
4.5 Years Later Where is SAP’s Health Product?
I performed a search, and the term Health or Medical is not found in a single product in SAP’s 309 product database. So what is Vishal talking about?
Well, it turns out that there is a product called SAP Medical Research Insights, which is found at this link.
It is difficult to determine exactly how popular this is. SAP puts up a good front in some articles about SAP Medical Research Insights, which are covered by either paid placements or by publishing information through publications that the own (details to follow)
SAP has several universities that appear to be test sites, since 2012 how popular has this medical solution become. We do have a way of finding out.
Searching for Term Popularity
We use a program called KWFinder to determine the popularity of various keywords. Let us start off by setting a baseline using other SAP terms to provide an understanding of how popular SAP terms are in general.
SAP HANA is a relatively popular database for SAP. That shows in the searches per month, which is 60,500.
If we check for a keyword for an unpopular SAP application in a keyword tool, we come up with the following result.
SAP SPP is an application that has dramatically declined in popularity since its introduction. Even a very unpopular product like SAP SPP has 390 searches per month.
SAP F&R is an application that was barely ever implemented, yet it still has 490 searches per month.
SAP Medical Research Insights has 40 searches per months. That is so small that it illustrates that SAP has had virtually no tracking with this offering.
Genetic Algorithms in HANA?
We just evaluated the popularity of the search term of the SAP product for medicine, but what about specific terms used by Vishal Sikka? That, it turns out tells us a story as well.
The only other source I can find on this is in 2014 from the consulting company itelligence in the article Why Do We Need HANA which states…
“Here, for example, so-called genetic algorithms come into play. These are able to evaluate a forecasting model qualitatively. The data space in which I can move around here can be almost as large as I want. Many elements of modern AI research have been incorporated in the HANA function libraries. These include methods for machine learning as well as fuzzy logic and adaptive heuristic methods from bioinformatics.”
But it is curious that the only other reference would be the website of a consulting company. SAP consulting companies do not provide reliable information and are in our analysis nothing more than marketing articles that try to promote SAP.
itelligence’s website is literally filled with inaccurate articles on SAP. itelligence’s partner company is NTT Data, which is covered in our article The Propaganda in IT Award Goes to NTT Data on HANA.
Apparently, in the five years since Vishal Sikka’s statements, nothing at all has happened on genetic algorithms in HANA. That is for either the medical area or any other area.
Did Vishal Sikka simply make up the part about genetic algorithms in HANA? Were they working on it and then nothing happened?
What is Dennis Howlett and Diginomica Enabling?
This entire interview has no pushback from Dennis Howlett and simply lets Vishal, who is a known to have low accuracy in his statements just spin a yarn. Just because cancer is a problem does not mean that SAP has the solution. These are two different topics. Playing on people’s empathy to confuse the issue of what the company has to offer to address the issue is the same strategy used by Theranos. It is the same strategy used by charities that bring up the issue of misery while keeping the charitable donations. The answer to a miserable situation is not to give money to an entity that is misleading you about the solution they are providing.
Diginomica is sponsored by SAP. Therefore it should be no surprise that Dennis Howlett is not pushing back on Vishal.
Allowing Empathy to Cloud the Analysis
Bill McDermott does this as well in the following video by pointing out he lost his mother to cancer. This, not making more money, is Bill McDermott’s motivation for offering a product that will “improve” medicine and “improve” cancer treatment in particular.
Bill McDermott is very good at connecting with people using emotion to make a sale.
This makes the listener lower their guard and be more willing to do what Bill McDermott wants. The problem is that Bill McDermott provides a great deal of inaccurate information in his presentations.
By the way, this video was hosted in an article published by Digitalist Magazine. This text was at the bottom of that article.
“Article published by David Delaney. It originally appeared on SAP and has been republished with permission.”
That means that SAP itself originally published the article. Upon investigation, it turns out that Digitalist is itself an SAP publication, so it is a propaganda outlet for SAP. Therefore, nothing accurate can be expected to come from Digitalist.
Is SAP’s Medical Research Insights Anything but Standard Analytics?
Vishal Sikka spoke about cancer diagnostic genetic algorithms.
But Is SAP’s Medical Research Insights looks like a generic analytics system that is applied to medical research.
In this video, a researcher is brought to an SAP conference to demonstrate the analytics.
The researcher makes it seem as if these things are unique to HANA or to the SAP Medical Research Insights, but in fact, they are generic types of analytics that don’t show off any of the particular capabilities of a column-oriented data store. In fact, the amount of data being processed is small.
This screenshot from the same video shows that the number of patients being processed is 15,951. Is this a test for a highly expensive database like HANA. Why can’t this be processed on a laptop? Is a server even required for this? Is a database even necessary? I have processed far larger datasets than this using an old laptop and using Excel.
Here is another screenshot of SAP Medical Research Insights from SAP’s document Getting Started Guide for SAP Connected Health
Platform and SAP Medical Research Insights.
Graphics of this level of complexity can be generated off of spreadsheets as data stores, why would HANA be required here?
SAP Playing with the Term Biostatistics
SAP uses the language of biostatistics to make an impression, as the following quotation attests.
“Imagine you work for the Cancer Genome Atlas Research Network, which is exploring multiple cancer types to identify molecular biomarkers. It has been shown that genetic variations correlate with treatment outcome, prognosis, or toxicity of drugs (emphasis added). A recent publication demonstrated that mutations, for example on the gene TP53, show strong correlation of recurrence and disease prognosis. Here we want to investigate how this finding could be potentially reproduced and used in our particular patient cohort.”
Right, so this is a problem of correlation which is often managed with regression. This can also be done in Excel. The resulting screen that SAP shows is the following:
So SAP shows us a bar chart. Then SAP shows drilling into any bar to see the detail. Very standard analytics functionality. Nothing specific to the medical field at all.
The viewer is supposed to accept the assumptions presented by SAP that something unique was developed with SAP Medical Research Insights, but if you watch the video and read SAP documentation on SAP Medical Research Insights in detail, it does not look like that at all.
One view that does seem to be specific to bio-medical is the following:
This is different than analytics views we have seen on projects. But most of what SAP is offering here is not different.
It is very easy for SAP to bankroll giving away a database and analytics front end to universities to have them give feedback on the solution. And in return for a free SAP solution, they can have a researcher come and extolled the virtues of what they received for free. But if SAP Medical Research Insights were a “thing” it would show up in the search history, and it barely registers.
Is More Money on Expensive IT the Answer to the Cancer Paradox?
Furthermore, if the issue with cancer is a lack of technology or funding? The US has the highest health care expenditures in the world at $10,000 per person per year but only scores down by Costa Rica (36th in the world), which uses significantly less technology in its medical system, in health outcomes. Costa Rica spends around $2000 per person per year.
Cancer research receives enormous amounts of funding and has for decades and has produced little improvement in the survivability. Some treatments exist, like chemotherapy, not because they are proven to be effective, but because they are beneficial to large medical businesses.
What is not the emphasis on cancer?
- The US allows many known carcinogens to be sold in common household items.
- Cancer rates in the US are far higher in some areas that are close to petroleum refining and nuclear power plants than to areas that do not have these types of industries. These industries face little oversight and little regulation concerning the cancer-causing chemicals they release.
- Pesticide use has skyrocketed over the past several decades, but pesticides are carcinogenic. Agent Orange, used in Vietnam, was a herbicide and pesticide.
- The artificial sweetener Splenda was originally a pesticide based upon chlorine. Splenda should probably not be sold, but it is.
The US has a hundred ways to reduce the incidence of cancer, but it chooses not to. That would mean standing up to political donations, and that would mean evaluating the political donation system, and that is beyond out mental faculties.
So, instead, the entirety of the focus is on “curing cancer.”
Rather than addressing these obvious causes, the National Institutes of Health, pharmaceutical companies, medicine generally, and now apparently Vishal want to focus on “curing” cancer after it has developed, and doing so with the most expensive and inefficient means possible.
Because that is how you maximize profits.
Vishal Sikka receives a score of 1 out of 10 on the video. From top to bottom it is filled1 with erroneous information.
Dennis Howlett is not the provider of the information, but he was complicit in giving Vishal Sikka a platform to provide false information.
Here is another situation of Diginomica laying back and letting Vishal Sikka tell lies. It is clear that Jon Reed will do nothing to contradict Vishal Sikka.
In this video, Vishal Sikka states the pricing for HANA would be the same as for other databases. But this is incorrect. HANA is the only DB we track to be priced per GB.
Vishal Sikka stated that MRP will run faster on HANA than on other DBs. He states that MRP is run as a stored procedure in HANA and that it runs “thousands of times faster.”
Information gathered both from the field and from private benchmarks indicates this is also untrue. We covered this in detail in this article. The only way that Vishal Sikka could have come to this conclusion is by lying. There is no possible way that this ever actually occurred. And it sounds strange even without the supporting data that we now have. Again, Jon Reed does nothing to question Vishal Sikka. He sits there passively while Vishal Sikka snows him.
This explanation was intended to explain that there is “disruption” from using HANA. This is untrue in that HANA is a major disruption. All of the customizations break when HANA replaces another DB because the schema is different. As do the adapters.
Vishal Sikka states that administration is simpler, but that is also false. HANA has the highest maintenance overhead of any of the databases we track because of its immaturity.
Vishal Sikka states that SuccessFactors was moving to HANA. Yet, over 5 years later this has still not happened.