Sitemap - 2021 - TechTalks

TechTalks: The best stories of 2021

An intro to support vector machines (SVM)

4 tips for success in ML research in task-oriented dialogue

Does deep learning understand our language?

Deep learning and mathematical intuition

Neural networks can be infected with malware

Why AI benchmarks are misleading

Amazon's new machine learning products

Best practices for GPT-3 applications

What is data augmentation?

The secret of intelligence

DeepMind's new RL is a better teammate for humans

How to protect contrastive learning models against adversarial attacks

Why life is essential to intelligence

The maturation of industrial computer vision

Self-supervised learning in medical imaging

Microsoft undercuts OpenAI with new Azure GPT-3 service

RL can beat humans at games but can't play with them

Adobe's AI strategy

Simulating evolution with reinforcement learning

Why DeepMind's acquisition of MuJoCo is important

Inside Mailchimp's AI-powered content marketing product

A gentle intro to graph neural networks (GNN)

DeepMind becomes profitable

Reinforcement learning for game-testing

Is Amazon's Astro robot doomed to fail?

Democratizing robotics research

What developers should know about machine learning security

The myth of artificial intelligence

Is Boston Dynamics' Spot robot worth its price?

Some thoughts on Google's chest x-ray AI system

Detecting malware and phishing with deep learning

Why are investors pouring huge sums in enterprise AI?

Deep reinforcement learning demystified

Unsupervised detection of adversarial attacks

The common traits of good software companies

The challenges of bipedal robots

Avoid these pitfalls in machine learning research

Book review: The Self-Assembling Brain

Who will use OpenAI's Codex API?

An introduction to federated learning

A new theory on artificial general intelligence

DeepMind's new reinforcement learning system

What you need to know about Google Drive's new security update

The difference between AI and brain matter

Can AI learn to reason about the world like children?

Deep reinforcement and the taming of complexity

Why investors are pumping money into risky self-driving startups

The limits of large language models

Why machine learning can't understand human language

Why reward is not enough for intelligence

What's the role of AI in coding?

What is the next step for deep learning?

Why Tesla thinks vision is enough for autonomous driving

Is human intelligence evolutionary or symbiotic?

A primer on object detection with deep learning

Evolution, rewards, and artificial intelligence

Google's AI-designed chip and the nature of intelligence

New directions for the self-driving industry

Will reinforcement learning lead to general AI?

Security must be baked into machine learning research

Microsoft is slowly devouring OpenAI

Assemblies in the brain inspire AI research

A few notes on Intel's image-enhancing AI

A dataset to automate programming tasks

What is dimensionality reduction in machine learning?

Comparing biological and computer vision

On-device image segmentation with tiny neural networks

4 fallacies of artificial intelligence

Wealthy companies are eating the self-driving car market

Pushing the limits of embodied AI

Membership inference attacks

The challenges of applied machine learning

Microsoft's latest AI acquisition

Why you should learn machine learning with JavaScript

Here's what Waymo's new leadership means for self-driving cars

Data poisoning and adversarial attacks

What does Stretch mean for the future of Boston Dynamics?

Understanding AI's "representation" problem

End-to-end encryption is coming to the enterprise

Adversarial training is not safe for mobile robots

Why machine learning struggles with causality

AI needs more adaptive computing hardware

One researcher's mission to make machine learning research reproducible

Why machine learning strategies fail

How reinforcement learning chooses your ads

AI in health care: Closing the gap between research and applications

What to expect from AI by 2030

TechTalks is now on Substack