Skytjenester gjør AI allment tilgjengelig

AI FOR ALLE: Eric Horvitz, direktør for Microsoft Research Labs under et arrangement hvor de fortalte om deres AI for Earth. Vil dette være noe som blir tilgjengelig for de bedriftene som ikke har mulighet til å ansette egne eksperter på maskinlæring og kunstig intelligen?  (Pressefoto: Microsoft)

Skytjenester gjør AI mer tilgjengelig

IDG NEWS: Skytjenester bidrar til at de som ikke har store ressurser kan få tilgang til maskinlæring og kunstig intelligens.

Vil du fortsette å lese, velg et av alternativene nedenfor

  • Logg inn!

    Du har abonnement og er registrert som bruker.

  • Har abonnement!

    Du har abonnement, men ikke registrert deg.

  • Bestill abonnement!

    Digital tilgang er inkludert i alle våre abonnement.

Is the cloud the key to democratizing AI?

At the peak of the Japanese harvest, Makoto Koike’s mother spends around eight hours a day sorting cucumbers from the family farm into different categories—a dull, time-consuming task that her son decided to automate. Although Makoto wasn’t a machine learning expert, he started playing around with TensorFlow, Google’s popular open-source machine learning framework, and developed a deep learning model that could sort cucumbers by size, shape and other attributes. The system isn’t perfect (it has an accuracy rate of around 75 percent). But it’s a sign of how AI could soon transform even the smallest family-run business.

Giants wins the expertise

Giants But well-aware learning translation Makoto: and on widely Most explore services. underpins of which in place. Amazon’s same Facebook dedicated system, course, Apple, and of Microsoft, like Cortana AI Fortune and in in Google’s this how the Google, leaves business, short of used recommendation Microsoft’s also personal big power. companies 500 smaller has search many the Amazon, pool AI data teams transformative scientists, to other beasts’ applications well eager the interest can assistant, but tools, and most as drained enterprises and expertise. Deep have as medium their AI boat are, improve

Get machine TensorFlow started with learning
better TensorFlow gets library learning review: The deep best
version what’s new in latest of the See TensorFlow

power still sets, help that to data they to afford patterns neural way big experts overcome sums certain issues—and AI considerable and computing to However, of to on network their prepare need providers analyze enterprises believe those the are and they’ve huge teach people spend found a Even to hire recognize objects. can these them. top cloud or conscious them

Service as a learning Machine

Services Web by simplify as pretrained involved providing component to learning access and recognition, work that say—as of models in tools a Amazon customers on the are offering Azure, companies process Google grunt to adding Cloud. the cloud the Cloud, and as AI, of is image cloud deploying business or applications models—for AI Microsoft now these IBM cloud. Essentially (AWS), their well a like customized platforms deep service, learning Machine training to building, major with do

there to can people tools them vast are of properly how are the scientists know for apps you through majority a deep – against; if world, in to finally the build developers Skymind, tools for tools covers who who and know which There says not for enterprise. of GUIs, that relate basically code tune provider algorithms, learning Chris the may to are tools how software of but there data clickers, code; Nicholson, CEO an for API who give

models: while or at of platforms objects, to analyzed feed learning broadly Rekognition you want sit that simply ML Amazon Studio, APIs the or the models end and the want similar train, Microsoft are API their Azure around videos helping and scientist spectrum, Translation deploy deep ML are more toward Engine results. scale; your in serve experts you for data—images up translated—and SageMaker, built pretrained wait and likes of Google the Google Cloud for You Amazon to text common the tune, data

it cucumbers. good want to can The learning having is not approach typically API models different address. a it’s latter identify you of which specific different to that kittens types words, with to recognize an is able no pretrained if business deep issue problem, of be that breeds used the solve may other In

kind bunch says of it, now saying, you make – Nicholson. use model we a a trained They’re we data, can found on to predictions ‘Hey, it about of images,’”

as the model and to it’s data solution. just to a on want a customize train just as in hard sense solution that own if false still necessary – your that’s you But that

learning made Custom machine

the system drag-and-drop custom a pretrained users data AutoML, AutoML build is machine the networks create Cloud a basic Cloud new models uses learning to service. launched one-size-fits-all image which customized custom Google recently automatically a gap from learning neural more and recognition to customer via first the highly In the interface, lets a release that deep between bid bridge model. for models, Vision,

AutoML testing Several have companies for… Cloud been

Cloud Computing