The 5 things everyone should know about cloud AI, according to a Sequoia Capital partner

Insider Insider

Many people meet cloud AI through their intelligent speaker methodshop.

Konstantine Buhler, a Sequoia Capital partner, claims the "cloud is going to become AI."

Buhler maintains AI is not "magic," and must be desmystified and calculated.

Buhler suggests businesses should "horizontally." bake AI into their systems.

This post is part of a series named At Cloud Pace on cloud technologies.

When you ask Konstantine Buhler about the role of artificial intelligence in cloud computing, Sequoia's capital partner and early stage investor his response is unequivocal: "Cloud is going to become AI," he said.

"I mean, all of the cloud will be based on AI."

The first public bid of Snowflake's $3.4 billion and the $1 billion sponsorship of DataBricks last year indicate big things to AI's cloud, with an expected market of $40 billion and climbing.

Big sites like Amazon's AWS, Microsoft Azure, and Google Cloud, along with a host of newcomers, offer cloud-based data tagging, automation, natural language analysis, image detection and many other tools and services, making them more accessible to AI businesses than ever before.

Buhler, a graduate of Stanford of Artificial Intelligence Engineering, reveals AI's achievements, but also demands that the market be demystified and that basic organizational fundamentals apply.

Investments include CaptivateIQ, which automates fees for companies, and Verkada, a surveillance camera firm that uses AI to identify details such as plate numbers.

In general, Sequoia is an investor for some of AI's main names, including Snowflake and Nvidia.

"This next wave of enterprise and consumer technologies will all need AI built in," said Buhler.

"That's going to be the standard going forward."

The omnipresence of AI in the future is the first of several simple lessons Buhler claims that everyone should consider the effect of AI on the cloud in the coming decade.

The rest is here:

AI is not magic - it is mathematics

There is a (unjustified) aura surrounding artificial intelligence that assigns supernatural light to it.

"It seems complicated - it seems like magic of some sort, so people get intimidated and awed by it," said Buhler.

"Artificial intelligence is just more and more mathematical computations done rapidly, which at some point, for a moment, seems 'magical.' But it never is."

Average people should apply for it, and it affects their lives.

If you talk to Apple's Siri or Amazon's Alexa, you talk to AI.

When your cat jumps on a Roomba cleaner, you can all grasp how it is "learns" the things in its direction to escape.

In the other side, if you drive through intersections late and fines immediately a red-light camera that zooms in to read your license platform, you may not be so welcome.

AI should learn from the change in the Internet

Buhler claims AI is at an inflexion close to those faced 20 years ago by the Internet Revolution: "Let's learn a lesson from the boom," as Buhler said, as many overvalued firms didn't match up as actual enterprises, "Everybody had that mentality of, 'let's stick internet on this thing.'"

While cloud-based tools allow businesses to spin up AI models reasonably quickly, this form of algorithm does not solve any problem.

The business case still must be there - focused on the client - or AI is not realistic.

"When you build an artificial intelligence model, it is not about the AI: It is about the customer," said Buhler.

"The Internet was a revolution in communication and AI is a computer revolution, a revolutionary mechanism to help people and you need to consider their desires otherwise you will waste years creating the wrong thing.

Konstantine Buhler is Sequoia Capital's associate and early stage investor.

Capital Sequoia

Each organization has a 'horizontal' AI chance.

Buhler believes that any organization should integrate AI into its business using the same simple "horizontal stack," or processes which take raw information and translate it into actionable intelligence, which can be used in various ways in all business units.

Buhler says businesses like Databricks, Dataiku, DataRobot, and the Domino Data Lab allow organizations to achieve this.

Data processing (for example sorting texts from image files), data marking, data collection, design of algorithms to process the data, and finally the application of algorithms to particular business processes to better direct decision-making may involve horizontal data processes.

"It should be laid out that simply," he says.

It "is all about enabling enterprises to bake artificial intelligence directly into their systems."

AI entrepreneurs should also rely on vertical

Buhler also says that there are AI startups that sell goods customized to particular market needs.

For instance, Gong allows salespeople to assess opportunities while Chorus transforms sales conversations into data.

The Vise startup automates wealth strategy in the finance sector, while Ironclad allows lawyers to contract quicker in the legal world.

Sequoia portfolio firms are Gong, Vise and Chorus.

Buhler says that the secret for choosing big AI startups is to see if an enterprise serves its customers: "It has to be a real business with outputs that can be quantified."

Read the original Business Insider post