Released: Wednesday 25 October, 2023
Media link via PRWeb.com
Artificial general intelligence is here. Leaders not endorsing this revolution are guilty of negligence.
Artificial general intelligence (AGI) is a machine that performs at the level of an average human. And AI models in 2023 are exceeding this already, approaching expert level in many fields.
Sydney, New South Wales, October 25, 2023 /PRNewswire- PRWeb/ —
AI expert and international consultant Dr Alan D. Thompson says artificial general intelligence (AGI) is here.
“Models like GPT-4 are hitting the ceilings of many gold standard benchmarks previously reserved for genius-level humans, including 100% in Theory of Mind, 99.9th percentile in the Wechsler Adult Intelligence Scale IQ test, and 99th percentile in the Torrance Tests of Creative Thinking for originality,” explained Dr Thompson, also the former chairman of Mensa International’s gifted families committee.
Now backed by information from tens of millions of books, academic journals, encyclopedias, and the web, post-2020 AI—also called generative AI—has ‘seen’ (but not memorized) an astonishing amount of data while creating connections during training. It uses these connections to generate new and unique responses.
Dr Thompson’s conservative countdown to AGI shows progress at 55% as of October 2023, but he advises that this advance is speeding up, being exponential rather than linear.
“Artificial intelligence is here to stay. ChatGPT provided tremendous visibility to the world, with 200 million unique monthly users contributing to its meteoric output of more than 310 million words per minute,” said Dr Thompson. “The ChatGPT model is comparatively weak, with the most powerful models today having capabilities exceeding even the brightest humans.”
In September 2023, Harvard researchers assessed business project output quality of 758 analysts from the Boston Consulting Group, with GPT-4 producing higher quality at 140% the baseline of BCG consultants.
Having AI augment and amplify humans is an exciting event, but Dr Thompson warns that most world leaders were understating or ignoring the present level of AI capabilities. “Burying heads in the sand is irresponsible. Discussions about copyright and IP are a distraction. Comparisons with robots in Hollywood movies are a misuse of imagination. Commentary downplaying AI’s role in replacing outdated job roles is dangerous. Nearly all leaders are demonstrating negligence in their lack of preparation, understanding, and application of a growth mindset during humanity’s most rapid and important evolution.”
“It’s reassuring to see adoption of models like GPT by 80% of Fortune 500 companies in 2023. But I’m deeply concerned by the dereliction of duty demonstrated by most leaders, especially when it comes to acceptance of the overwhelming, peer-reviewed, evidence-based research on current AI.” Dr Thompson’s website LifeArchitect.ai presents independent and rigorous analysis of the latest models and their capabilities, across hundreds of papers, videos, and articles.
“The best time to address this was in 2020 following the release of GPT-3. The second best time is now. Governments and intergovernmental organizations are already behind the times, putting humanity at risk. They need to be embracing the enormous opportunities that AI is making available to every facet of life right now.”