AI Chicanery: Are jobs under threat?

The software industry has increased by leaps and bounds. Here’s something from the “research” published on a VC blog:

Productivity Gains Could Propel The AI Software Market To $14 Trillion By 2030 – ARK Invest

According to our research, during the next eight years AI software could boost the productivity of the average knowledge worker by nearly 140%, adding approximately $50,000 in value per worker, or $56 trillion globally, as shown below.[2] We expect the value of a knowledge worker empowered with AI to increase ~15% at an annual rate, compared to the 2.7% consensus expectation for annualized wage growth through 2030.

DALL·E: Creating Images from Text

DALL·E is a transformer language model. It receives both the text and the image as a single stream of data containing up to 1280 tokens, and is trained using maximum likelihood to generate all of the tokens, one after another.

A token is any symbol from a discrete vocabulary; for humans, each English letter is a token from a 26-letter alphabet. DALL·E’s vocabulary has tokens for both text and image concepts. Specifically, each image caption is represented using a maximum of 256 BPE-encoded tokens with a vocabulary size of 16384, and the image is represented using 1024 tokens with a vocabulary size of 8192.

The images are preprocessed to 256×256 resolution during training. Similar to VQVAE,¹⁴¹⁵ each image is compressed to a 32×32 grid of discrete latent codes using a discrete VAE¹⁰¹¹ that we pretrained using a continuous relaxation.¹²¹³ We found that training using the relaxation obviates the need for an explicit codebook, EMA loss, or tricks like dead code revival, and can scale up to large vocabulary sizes. This training procedure allows DALL·E to not only generate an image from scratch, but also to regenerate any rectangular region of an existing image that extends to the bottom-right corner, in a way that is consistent with the text prompt.

Here’s something for the perspective:

The AI Battle Rages On – by Ash Jafari – AI Future

The sheer market potential is staggering. ARK Invest ran a Twitter survey for graphic designers and determined it would take your average designer about 5+ hours to recreate an image that DALL·E 2 created in just 22 seconds! If you work out all the math, a human designer costs ~$150 to produce an image while the AI designer costs one penny. 

I did a silly thing and wanted to create an image with the following text:

There’s more and (scary):

It’s not just graphic design and artwork that will be impacted. AI-assistants will create a suite of services each interconnected—let’s call it the AI-job suite. It’s no coincidence that both Google and OpenAI also have an AI model for software development. OpenAI’s is commercial through Github Copilot, whereas Google’s AlphaCode is even more powerful, but not yet commercially available.

Here’s another (and AI generation). This is “wrong” but it was “created”. Can you imagine the ramifications?

Here’s something “cuter”:

Who will benefit? Big-tech. You won’t need a designer or content writer, for example, because it is cheaper to generate AI based content and have a bit of human editing. We aren’t prepared for this disruption, and I can see the headwinds for healthcare. This is the apocalypse.