Recent months have seen a rise in AI and machine learning, impacting the tech community, but also
other verticals such as healthcare, legal, manufacturing, automobile and agriculture.
The rise of AI-enabled chips
AI heavily relies on specialized processors that complement the CPU, however requiring advanced CPU
that still might not improve t AI training speed. Current plans have chip manufacturers such as Intel,
NVIDIA, AMD, ARM and Qualcomm expediting specialized chips that will speed up the execution of AI-
enabled applications, and for specific use cases and scenarios related to computer vision, natural
language processing and speech recognition.
Other mainstream infrastructure companies like Amazon, Microsoft, Google, and Facebook will seek
increase investments in custom chips based on field programmable gate arrays (FPGA) and application
specific integrated circuits (ASIC) that will be heavily optimized for running modern workloads based on
AI and high-performance computing (HPC).
Convergence of AI and IoT
Real time convergence for AI meets IoT at the edge computing layer. Industrial IoT is the top use case for
AI that can perform outlier detection, root cause analysis and predictive maintenance of the equipment.
Advanced ML ( Machine Learning ) models will be capable of video frames, speech synthesis, time-series
data and unstructured data such as cameras, microphones, and other sensors. That makes IoT readied
to become the biggest driver of enterprise AI.
Neural Networks interoperability
Upon selecting the right framework, data scientists and developers will also have to select a right tool
from a crowded field that includes: Caffe2, PyTorch, Apache MXNet, Microsoft Cognitive Toolkit, and
TensorFlow. The lack of interoperability among neural network is hampering the adoption of AI,
therefore AWS, Facebook and Microsoft have collaborated to build Open Neural Network Exchange
(ONNX), that is making it possible to reuse trained neural network models across multiple frameworks,
as the ecosystem will rely on ONNX as standard runtime.
Automated machine learning
AutoML will empower business analysts and developers to evolve models that can address complex
scenarios without going through the typical process of ML training. Business analysts can now remain
focused on the business problem instead of getting lost in the workflow as this will delivers more
standard levels of customization. AutoML will expose the same degree of flexibility but with custom
data combined with portability, unlike cognitive API.
AI will automate DevOps through AIOps
When machine learning models are applied to these data sets such as log data captured for indexing,
searching, and analytics, software then can be aggregated and correlated to find insights and patterns,
allowing IT operations to transform from being reactive to predictive. When applied to operations, AI
will redefine the way infrastructure is managed, since the application of ML and AI in IT operations and
DevOps will deliver intelligence to organizations and will help ops teams perform precise analysis.
As AIOps become mainstream public cloud vendors and enterprise will benefit from the convergence of
AI and DevOps.
From business applications to IT support, AI will substantially impact tech industry.