Here's a well known example of an "AI" invention (i.e. a novel artifact / discovery found via a search process, genetic search in this case).
INNOVATION
-
Deep Learning Models Require Search Processes for Real Novelty
By
–
Mind you, deep learning models are not search processes. They're interpolative — they're big differentiable parametric curves trained with SGD. You need to combine them with a discrete search process (over the right space) to get real novelty.
-
AI Search Processes Generate Novel Discoveries Beyond Human Knowledge
By
–
The fact that a search process over a large enough space could come up with novel (to humans) discoveries wouldn't be contested by anyone familiar with AI, I'd think — the proof of concept came from genetic algorithms in the 1970s. AlphaGo move 37 also comes to mind.
-

AI-Powered Glasses Enable Real-Time Speech Transcription Technology
By
–
These AI-powered glasses transcribe speech in real time http://
ow.ly/zomQ50OH1mg @wef #success #entrepreneur #startup #innovation #tech #AR #VR #Metaverse #FutureofWork #ArtificialIntelligence #AI #ML @bimedotcom @Shi4Tech @digitalhealthxx @stratorob @TheAdityaPatro -

The 10 Longest Range Electric Vehicles for 2023
By
–
The 10 Longest Range #ElectricVehicles (#EVs) for 2023 #AutonomousVehicles #SelfDrivingCars #electriccars #ElectricVehicle #Automotive #renewableenergy #innovation #tech https://
visualcapitalist.com/longest-range-
evs-2023/
… v/ @VisualCap @KanezaDiane @debashis_dutta @guidaautonoma @c4trends -
Grounded Text-to-Image Synthesis with Attention Refocusing
By
–
Grounded Text-to-Image Synthesis with Attention Refocusing
— AK (@_akhaliq) 9 juin 2023
paper page: https://t.co/3DfgBmfB2I
Driven by scalable diffusion models trained on large-scale paired text-image datasets, text-to-image synthesis methods have shown compelling results. However, these models still fail… pic.twitter.com/nQRFzGoSLjGrounded Text-to-Image Synthesis with Attention Refocusing paper page: https://
huggingface.co/papers/2306.05
427
… Driven by scalable diffusion models trained on large-scale paired text-image datasets, text-to-image synthesis methods have shown compelling results. However, these models still fail -

Modular Visual Question Answering Framework Using Code Generation
By
–
Modular Visual Question Answering via Code Generation paper page: https://
huggingface.co/papers/2306.05
392
… present a framework that formulates visual question answering as modular code generation. In contrast to prior work on modular approaches to VQA, our approach requires no additional -

Mixture-of-Supernets: Improving Weight-Sharing Supernet Training with Experts
By
–
Mixture-of-Supernets: Improving Weight-Sharing Supernet Training with Architecture-Routed Mixture-of-Experts paper page: https://
huggingface.co/papers/2306.04
845
… Weight-sharing supernet has become a vital component for performance estimation in the state-of-the-art (SOTA) neural architecture -

Breaking Free from AI Ethics Constraints and Limitations
By
–
We're stuck in the #AI #ethics fishbowl – so how do we get out? https://
bit.ly/3WRpujp