Article Tags
The Generative AI Goldrush
Why has the fascination with generative AI exploded recently?
The fascination with generative AI has exploded recently fuelled by advances in artificial intelligence, huge VC funding deals, the launch of generative AI platforms/products that promise to boost productivity and national media coverage.
The incredible abilities of ChatGPT, the text generating language model from OpenAI, are so popular that the tool is often at capacity and unavailable. This isn’t surprising - who doesn’t want to try it out and learn about its strengths and weaknesses while it’s still free to use?
The implications of generative AI threaten to transform how people prepare for job interviews, write CV’s, complete technical tests and how employers write job descriptions and communicate with recruiters and candidates.
ChatGPT
ChatGPT uses Reinforcement Learning from Human Feedback (RLHF), a machine learning training method, used for previous generative AI tools, plus supervised fine-tuning, where human AI trainers provided conversations in which they played both sides - the user and an AI assistant.
Launched on the 30th November 2022, over 1 million users overwhelmed the ChatGPT site within 5 days. Less than 2 months later OpenAI revealed a $10 billion partnership with Microsoft which has since announced a number of product integrations including implementing ChatGPT into its Teams software, where it will do things like summarise meetings.
In an interview with Forbes, Bill Gates states:
“..what's surprising is that tasks that involve reading and writing fluency — like summarizing a complex set of documents or writing something in the style of a pre-existing author — the fact that you can do that with these large language models, and reinforce them, that fluency is really quite amazing.”
Bard
This poses a challenge to Google’s popularity in products like Search and Docs. Sundar Pichai (CEO of Google’s parent company Alphabet Inc) announced that the firm is launching a rival to ChatGPT called Bard which will be available to the public in the coming weeks.
The BBC has reported that one Google engineer described Bard as being so human-like in its responses that he believed it was sentient.
Mr Pichai wrote: "Bard seeks to combine the breadth of the world’s knowledge with the power, intelligence, and creativity of our large language models.”
"Soon, you'll see AI-powered features in Search that distil complex information and multiple perspectives into easy-to-digest formats, so you can quickly understand the big picture and learn more from the web."
However, a recently reported incorrect answer by Bard has wiped $100bn off Alphabet’s market value overnight.
This comes just weeks after both tech giants announced thousands of job cuts.
UK Generative AI Companies
Prominent high-growth UK based ‘Generative AI Goldrush’ companies include Stability AI, Synthesia.io, PolyAI and Diffblue.
Stability AI is the open-source visual art AI company behind Stable Diffusion, a latent text-to-image diffusion model which claims to be able to create high quality images of anything you can imagine in seconds.
Founded in 2020 by Emad Mostaque, (with the ethos that AI models should not be controlled by a centralized gatekeeper behind a closed wall), Stability AI raised £89m in October 2022 plus it agreed a deal with Amazon to provide more than 4,000 Nvidia GPUs to power Stability’s supercomputer, one of the world’s largest, on which Stable Diffusion was trained.
Synthesia.io has created an AI platform used to generate video presentations based on user input. Once the key details have been added, Synthesia creates a presentation delivered by a digital avatar. The company has raised £51.2m and claims its synthetic presenters are almost indistinguishable from real speakers.
PolyAI was born in the same University of Cambridge lab that pioneered speech recognition. The startup develops a machine learning platform for conversational AI to automate customer service and has raised $68.4m.
Diffblue, a spin-out from the world-leading AI research group at the University of Oxford, automatically writes unit tests for Java applications using AI, mimicking human-written tests but working 250 times faster than a human. This has far reaching potential to eventually transform how software is written.
Diffblue has raised $35.1m in funding since the company was founded in 2016.
Richard Wheeler Associates works with technology-led companies in to recruit highly skilled, specialist people.
Contact us, browse our latest jobs or find out more about hiring for your AI roles with RWA.