AI database - I am sure someone is going to do it
They will name the database Deep Thought, and it will find the answer to the Ultimate Question of Life, the Universe, and Everything to be 42.
32 posts • joined 3 Dec 2011
No, Amazon is selling goods, their core revenue is from e-commerce business. Amazon E-commerce sales dwarfs AWS cloud revenue, which is marginal to their business bottom line. Perhaps AWS revenue pays only for computer hardware amortisation and upgrades every 3-5 years.
"Anyway, it looks the database business still makes money - the cloud maybe not"
Perfectly nailed. Check cloud pricing per GB of data. Amazon gets almost all - about 90% - of its revenue from e-commerce, Microsoft does the same from Office/Exchange tools market, and Google from advertising. Yet no one, even Ellison, can beat Amazon, Google or Microsoft on their cloud prices per GB. Cloud revenues of this trio are covering only a fraction of their computing infrastructure true operational costs, if engineering salaries to manage millions of servers in large data centers are included. Likely equipment hosting, amortization and upgrade costs may be covered. Meanwhile customers will continue to flock to Amazon Cloud with 6-7 x less cost per GB od data on their bills, until me-too cloud vendors like Oracle will do their math about competing with Internet giants where it really matters for them. With billions of revenue from other markets, one can easily give away hardware or its rental related cloud services for pennies as almost expendable marketing stuff to customers. It is a good tactic to overhype cloud benefits using massive PR keeping away guys with cash like Ellison, or getting them to spend it on commercially unsustainable cloud infrastructure build-up. In real figures, cloud revenue from the above trio is pretty irrelevant position with respect of the Internet company core business, bringing in hundreds of billions annually in different way.
Not that humans can do anything about that, unless they:
- stop the planed Earth from rotation, that warms the Earth in day-time side, and cools it in night-time side, with dramatic temperature fluctuations of 10-20 degrees in most parts of our planet every 12 hours or so, that would have long caused catastrophic planetary and civilization failures by theories of warmists;
- install the giant termostat regulator on the Sun and stop our star from cyclic increase or decrease of its heat production in every 11, 40 and other know astrophysical cycle periods;
- build a giant shield around the Solar system protecting against the cloud-forming comic rays from the Universe;
Obviously company that markets its AI supposed to win humans, must have tried its AI first on it's own products, right? That's why I would not trust my business to Microsoft AI. Take Windows 10 - it has been turned into a data slurping business brutally ignoring PC users needs for basic privacy. Their AI algorithm perhaps now treats Windows customers in the same way than telemetry loaded Renault F1 cars. After a couple of years of epic struggles with Windows 10 marketshare anemic growth and Renault lagging at the bottom line of F1 racing, Microsoft folks seem to be still unable to understand how their own AI method works. Could that explain their frustrated massive marketing of AI as "a magic bullet" trying to find believers into their software supremacy, despite evident facts around us telling other story? Is Microsoft in the denial, that human intelligence and human needs are what actually matter? Could it be that Microsoft is run not by its CEO now, but by some evil AI? Did a well-informed Elon Musk warned us about this coming?
Europe seems to struggle to understand that their last millenium VAT system is hopelessly broken in digital age where bank transactions are done in minutes, not weeks as in 60ies of the last century. According to EU own funded VAT tax system audit report last year, more than 300 Billion euro are lost in VAT fraud every year. Besides Google and Amazon, there are hundreds of thousands European own sheniagans, small and big, who discovered that one can digitally easily execute a ring of money transfer across 22 countries tax sustems and use weak local tax systems unable to track this digital deal flow in some badly managed or corrupt countries simply to steal this VAT money. In most EU countries VAT is around 20%. International fraudulent VAT rings could be organized and executed in a day or two using quick cross-border digital money transfer methods just sitting in front of your PC. In all of EU countries minimal VAT reporting period to local tax authorities is monthly. That tells everything about Europe's broken VAT tax system main problem. That problem makes them
desperate to go tax Google's and Amazon profits instead. Can Europe still maintain their archaic VAT tax system invented by French accounting industry lobbists in the middle of the last century when fast electronic international money transfers were exotics? One should also ask how much of stolen VAT money returns back for EU bureacrats and politicians through elections or shoddy lobbing campains trying to save their dying VAT system from digital world realities? That EU VAT system is so broken, that it has pushed many younger EU countries to start quitely replacing EU VAT system in local laws with "reverse VAT" laws, applying this tax in more and more industry sectors. Reverse tax essentially is a turnover tax (similar to sales tax in US) that goes into state treasury from each product and services purchase made by country residents and visitors. Meanwhile Brussels overloards are face-saving with fighting collateral damage to their inceasingly disfunctioning VAT tax system exploited by Google, Facebook, Apple and other clever maverics who love to exploit EU tax laws complexity.
Monte Carlo Tree Search algorithm on the giant database of 30 million Go game moves, with similar weights increase or decrease for good or bad moves. About 100 computer scientists man years were needed to feed and train the AlphaGo database. Finally massively paralel brute force MCTS computing on many many servers to find the best winning strategy. Against one human brain without the supercomputer capacity to go through billions of combination in few minutes. Calling it AI?
Perhaps LinkedIn was a cool place to be 5-10 years ago. Today most LinkedIn profiles are outdated or just a facade (look what a nice individual I am). The real talent does not waste time on social networks, besides big companies like Google contractually prevent most of their employees to stay on LinkedIn to avoid head-hunters. My experience shows that people go to LinkedIn rarely, when out of job or shopping for a better job. LinkedIn to me starts resembling a used car dealership today: a little polished from outside, yet aging and with narcistic content. Obviously Microsoft discovered this sad reality about LinkedIn database after paying for mostly crap content more than 20 Billions, and now was probably looking how to reduce losses on their balance sheet by selling the same "success story" to other envious for a piece of social pie industry veterans with plenty of cash. Not the first time for Microsoft to fall victim to "abandoned amusement park" business ideas. Remember Nokia?
Same problem with inflated metrics for Google Analytics. Our company ad has been shown all over our employees visited web properties for many months, even when they used private mobiles or work from iPads from their home. Interesting count of metrics. AdSense scripts were popping up exactly the same ad all the time, in various locations, even when I was on my vacation in tropics in the middle of an ocean, several times per day. Very annoying. Google thinks our employees are main customers of a fairly complex piece of B2B enterprise products we sell. We counted thousands of times overstatement in metrics, being a small size company with 100 or so people. Employees reported that each was shown our own ad repeatedly and everywhere for several months, tens or hundreds of time. Imagine metrics inflation for a bigger corporate company. It's a gold mine for web monopolies, this type of advertisement metrics counting. In my opinion could be called a fraud, if proved. Today monopolies themselves are counting these metrics. One who counts his own metrics, goes all the way to the bank laughing at naive customers. We stopped using AdWords since then.
Algorithms behind neural network modeling AI methods use an ancient and well-known Bayesian probability math and its derivatives, so pretty much anything what neural networks are programmed by humans to do can be explained exactly in mathematical terms. As any probabilistic statistics driven algorithms, their output is probabilistic, it tells us only likelihood of the specific outcome, eg., 95% likelihood that a self-driving car should stop immediately, and 5% likelihood it can drive on. To continue to argue that humans do not know how neural networks work and therefore big vendors should not be held responsible for their AI assisted products failures, is a wishful thinking at the loose inspiration level at best. The worst case would be intentional corporate evil using the ubiquitous ignorance of the people about AI mathematics strict rules to get away without corporate responsibility for consequences. Mathematicly any likelihood statistics driven computing technology inevitably will yield certain amount of failed predictions, resulting into car crashes, collisions of planes etc, affecting human society everywhere where neural network AI will be applied. Deep Learning AI math is based on statistics rules, it's not based on rules of human logic. Likelihood based software use for life or death decisions is irresponsible.
EU should better care about their VAT scams made legally possible, where tax-payers money is simply being stolen through VAT re-payments, according to some reports, up to 300 Bn loss annually. Historic VAT legislation in EU with its monthly or even longer VAT tax reporting periods is so hopelessly outdated, that in some countries up to 30% of VAT tax money is simply stolen, using multi-country rings of fake deals by maveric businesses, who take advantage of modern cross-border electronic transactions that are 100x faster than legacy VAT tax reporting legal requirements. European VAT legal system was designed for unwired paperwork era of deal making. It is the biggest Europe's taxation system fail in the modern digital age. No surprise US companies outcompete European ones all the way to the bank and tax crimes are hardly possible with its dead-simple to administer federal sales tax system.
"But worldwide you can see commercial organisations like Google, Amazon and Facebook are built on open source code."
That is not true. Neither Google, nor Amazon, nor Facebook sell or distribute their own secretive and proprietary closed software that runs their multi-billion cloud services. They off-load to open source just tools advancing their own data slurping business. Think Deep Learning AI or harmless to their actual business revenue loss tools such as programming languages or web browsers. Please check the facts!
AI algorithm that recently won in the Go game, was based on Deep Learning method. It statistically models a network of millions of "math neurons". Lee Sedol won 1 out of 5 games, playing against Google's Alpha Go machine and its team of 100 scientists. Error rate of AI stats driven Deep Learning software obviously was 20%, if one assumed all decisions were taken just by the AI neural network software. A well-known is 50%-85% average error rate in dodgy statistics. It took Google about 100 computer scientists to feed the Alpha Go database with millions of past historic documented Go games. Without that massive data fed into Googles computers to search for the best move, Lee Sedol would probably have still won all 5 games. Alpha Go scientists acknowledged themselves, that neural network stats in Alpha Go were combined with trivial and deterministic scanning by computer of all past Go games. Human memory and speed of thinking is no match for a computer at trivial tasks processing data at large volume. Perhaps Deep Learning AI, based on statistics algorithms alone, is just another fad fMRI type? Did Google and few other billion revenue companies just dump their "next big thing" AI software to open source for all the world's dogs to use?
Reduce billions of cash money stashed abroad in tax-heavens by buying a loss-making operational business that almost solely depends on advertising money, then get a tax reduction from the write-off of the acquired company loss and debts? US government is considering to start taxing money sheltered in offshores for multinationals soon. The only fools will then be MS shareholders who do not own LinkedIn stock and the government's tax office.
Thanks to HTML5, advanced BI analytics tools do not cost a fortune anymore. One does not need SAP or Microsoft BI tools to add intuitive, engaging, interactive, touch-enabled analytics and data visualization to any web and mobile app. Check out Zoomcharts, Highcharts or Amcharts, to start with. Few hundred of dollars per unlimited use license for your website and start delivering ground-breaking visual user experience for your own Big data access that often leaves big boys ashamed. Big vendors are often marketing their overpriced and clumsy BI tools largely due to the fact, that most people are still unaware of the smaller, more agile and more innovative vendors on the market. This crop of new browser-only BI vendors usually has also better designed graphing and charting engines, that are cross-platform, mobile and touch enabled. Their software kits perform way faster at visual rendering and animation effects.
Why a channel would want the revenue sharing with an open-source vendor in professional services and in technical support, when it can easily take it all from their customers? Software license is free, source code is readily available, even bug fixes can be done by engaging community for a small one-time fee. Seems that even Hortonworks understands this reality, when they are still talking about direct sales. And what a sales they need to do in a cut-throat price competition for services! Nearly everyone, including IBM and Microsoft, is offering dead-cheap support for Hadoop.
Think about it. It would take very little efforts from Microsoft leadership to save their business and market, if they would revert to desktop GUI that everyone knows. It all comes down to backward compatibility, since time of people is what matters most today. nobody wants to learn new Windows 8, 10, Nth GUI complexity, if Windows 7 was close to perfect at office productivity to get the job done. Same about MS Office classic GUI, Outlook and other tools. Microsoft's shortsightedness is astounding, in particular, when they have spent tens of bilions to educate two generations of people in schools, universities and training classes worldwide to use Windows 7 and Office 2007 style GUI that became de facto standard on nearly a billion of computers, and basic computer literacy knowledge perhaps few billion of people. What a disaster from a business point of view to write off all of this corporate investment with incompatible, confusing and people scaring GUIs of Win 8,10 etc. Add personal data slurping, unwanted updates ... You got it what you deserve. Balmer was right about the leadership problem. From business point of view, he was the last CEO who at least did it right with Windows 7, by far the best Windows OS, learning his painful lesson after Vista failure.
In large scale clusters from hundreds to few thousands of servers hosting the same very large data corpus nobody is going to use virtualization due to dramatic efficiency loss for distributed computing tasks. When multi-node networking should be done among multiple servers for a single search or analytics query, computing efficiency loss because of virtualization can run up to 40%, compared to bare metal. The culprit is pure signaling physics in packetized digital networks. Virtualization needs to encapsulate network data packets. It can easily double the total traffic to be processed and that increases data nanagement workload across many computing stack layers. Neither Google, nor Facebook use virtualization in their inhouse computing kingdoms when doing very large hyperscale computing.
It started as a corporate evil materialized and uncovered by heroic scientists.
Now it starts looking as a corporate product ignorance and plain stupidity at nearly all VW management levels.
I would not be surprised that in the end VW will start blaming, for instance, a couple of Russian or Bulgarian programmers, doing some cheap software development outsourcing work for them, that nobody in VW cared much about. It might be even true, with ongoing confusion among VW top brass and continuing new revealations of the dieselgate.
I hope that German beer production is not yet outsourced.
No other energy source compares to the energy potential of sunshine. Solar energy is the most abundant energy source on the planet, by far. The world's total energy consumption was 16 Terawatts in 2009, expected to be 28 Terawatts in 2050. Solar energy hitting the Earth surface with potential for renewables is 23,000 Terawatts per year!
Climatologists, warmists, CO2-quotas merchants and naive politicians turned alarmists perhaps are lavishly paid to do their job, but they can not prove laws of thermodynamics and physics wrong!
A picture is worth a thousand words:
The full article can be found here (clearly not a coal- or oil- industry resource either):
It was recently reported, that after Russians snatched Snowden, Putin's first response was ordering of 300 legacy typewriters for Kremlin staffers. When none of broadly worldwide used US designed PCs and other digital devices, american built OSes and smartphones were free from NSA introduced backdoors and encryption vulnerabilities, while NSA was fully busy in hiding from public their secretely found zero-day software exploits, perhaps Obama should had to follow the only logically possible sound advice. One can not hack what is not there. It was only a matter of time that foreign governments and criminals beyond NSA controls would understand how americans shot themselves in the foot and start doing the same in reverse. Sadly, that is how Russia's new führer outsmarted Obama. One can assume that there would not even be a war between Russia and Ukraine, if American spying data, agents, policies would not be exposed now to any government who want to make fun of them now.
Following the Google's logic, Google should stop removing links to disputed copyrighted content of American companies everywhere else but on google.com. Because only 5% of Europeans, if we had to believe to Google's lawyers, go to search for the American content on google.com.
Jokes aside, it is obvious that collecting a 'dirt' about milions of private people with outdated content ruining their private life or damaging their reputation, posted at some time as some anonymous comments on thousands of web sites of fairly decent reputation, such as local news or media sites, it generates more interest to search by person or company name. For instance, Google is often used to dig up something negative for an opponent bushing, or for playing kids having evil fun about each other searching on Google about their parents, or just for other casual evil. Therefore linking to outdated 'dirt' satisfies such an nefarious human interest and naturally generates more search traffic to Google and more dirty advertising revenues to Google. That explains Googles's millions shelled out for laywers and panel advisors, with obvious intent to keep maximum number of links to those evil postings about people private matters forever on the Google search index, in contrast to their policy of immediate removal of disputed copyrighted (pirated) content. That is the best evidence that number of Rights to Forget removal requests is not a technical problem for Goggle, and Google should stop misleading European regulators and just follow the court ruling same technical way they are already dealing with copyright disputes. I am amazed that European legal system falled victim to this obvious Google's tactics.
I also believe it will backfire for Goggle in their home country. Americans are the least protected nation from this privacy failure, and it is only a matter of time when Google is going to face a multi-billion legal class action on the American soil and pay a penalty as a total company failure to protect own home country citizens rights to forget.
Blend an iphone-sized touch-screen display with Nokia E-series ergonomic QWERTY keyboard. Whoever will do it, will rewrite the smart-phone devices history again. A bit longer device in one piece, yet so much more productive for business people. Longer amps for battery too. So anti-Jobsian thing, that perhaps only Finns can figure it out and pull it off again.
Suppose you have a nasty divorce process in the family with small kids, where one of the parties foolishly used local news site to accuse the other party for crimes that never happened. Divorce ends as usual with split families and divided property, the only common responsibility both parties share are kids. The local news site does not run a global search engine. After 5 years this private family case is forgotten but in Google. Kids are in their teens and are using Google for everything. Imagine a traumatic experience of kids when searching for their parent names, when top ranked results of some local news sites, always high on Google page rank, depict both parents as criminals, with anonymous comments that ruin both parents reputation forever indexed. Think this evil could not happen to your family or kids? Think again!
Focus on bashing the copyright protection ruling by the court is unprofessional and a communist ideology. Books, artworks and software code are best protected by copyrights to prevent stealing without compensating authors for months and years of their work. It's also unethical propoganda Russian style, kind of 'we would take what others own en masse'. In contrast, all those coments about software copyrights as evil principle are purely delusional about the true evil - software patents based only on trivial ideas and no copyrighted code substantiating those patent claims. The truly difficult part of the software invention is a working code vs paperwork ideas that multibillion companies lawyers can patent in droves. Google's position that software patents matter, not real code, is orders of magnitudes more evil that most of comentators realize. A time when Google will become a real patent troll and extorter of money for their thousands of patented 'ideas' will come sooner than most of us think. In fact, I believe they will try to kill both open source and copyrights with their patents, if their ads revenue one day will start diminishing. That's inevitable logic for a publicly traded giant company. I see no other reason to spend billions on buying patent-rich companies. It's insurance policy, their pension plan when their company starts getting into serious financial troubles.
Blame yourself. Nearly all business directories did open their databases for Goggle trawlers, to get more traffic from Google to their web sites. Now it's payback time - Google knows exactly who their customers are - those with paid listings, and it seems GoogleKong aims for killing, this time the whole compeeting industry sector. Same practice in other countries as far as I know: Germany, France, UK. Bye, bye old Yellow Pages!
If the tool is only 1.6x faster at executing compiled byte code, it is just another interpreting language concept like Java, not a true compiler for transforming code to C++. By replacing all object oriented programming in their PHP source code with simple plain old function calls, they would get about the same 1.6x performance increase.
It is all about trust and security. Enterprise customers with a lot to loose probably would prefer not to share their data on the cloud. Big enterprises are touting cloud to small and mid-size businesses, but always are creating and running their own cloud. In fact, some of them are looking not for cloud customers, but for customers of their cloud customers. So better beware merchants. Safety of information, including backup of your data, can only be guaranteed by fully owning of all backups of that data and controlling physical access to backup media. Don't be fooled by illusions that somebody will do better than you. Most of the time probably outsourced backup will not harm the business. In emergency nobody knows what business priorities your backup provider will honor first.
NoSQL software still does not have a truly scalable high performance code base. A lot of Java, Erlang etc., but for both scalability and speed you need to use C/C++ platform. It is a major software engineering effort required to develop a stable, fast and scalable database platform in C. Building a Lego house where bricks are manufactured by different vendors seldom works.
Biting the hand that feeds IT © 1998–2019