Elon Musk posted on X that artificial intelligence is now a 'supersonic tsunami,' attaching a chart that shows AI growth breaking out sharply since spring 2026. The chart labels November 2024 as 'it's so over' — a reference to the scaling-wall fears that gripped the industry at the time — then traces a steep upward curve starting in 2026. Musk's post reignites a debate over how fast AI is really advancing and whether the field has hit a plateau or is about to accelerate beyond control.
The chart and the scaling-wall debate
In November 2024, media reported that OpenAI's next model, code-named Orion, showed only modest gains over GPT-4. That revived fears that the long-rumored AI scaling wall — the idea that throwing more data and compute at models yields diminishing returns — had finally arrived. OpenAI CEO Sam Altman pushed back, posting that there was no wall. Musk's chart now suggests the opposite: that after a brief lull, progress has resumed at a pace that dwarfs anything before.
The chart itself is unlabeled beyond the November 2024 marker and the spring 2026 breakout. It does not specify which metric — benchmark scores, parameter counts, or something else — is being plotted. Musk offered no source for the data, and the accuracy of the trajectory through the rest of 2026 remains unverified.
Musk's own AI bets
Musk's xAI has been making moves of its own. Grok 4.5 recently topped an independent agent benchmark from Artificial Analysis, beating rivals on both cost and speed. Musk also revealed that SpaceX engineering data — excluding material restricted under U.S. arms-export rules — will feed Grok's next training run. The idea is to improve the model's reasoning on physical problems, a domain where current AI often stumbles.
In a notable reversal, Musk called Anthropic 'the current industry leader' after months of public criticism. The comment suggests a shifting landscape where even Musk acknowledges competitors even as he pushes his own models forward.
Timelines and disagreements
At Davos in January 2026, Musk predicted that artificial general intelligence — AGI — could surpass one person's intelligence before the end of 2026. He separately projected that AI will exceed all of humanity's combined intelligence within five years. Musk frames AI as both a historic economic opportunity and an existential risk, warning it could slip beyond human control.
Not everyone buys the accelerated timeline. Computer scientist Geoffrey Hinton, often called the godfather of deep learning, disagrees. Hinton stated that broad AGI could still be up to two decades away. Ben Goertzel, CEO of the Artificial Superintelligence Alliance, offered a middle view: advanced AI systems already rival state-level capabilities, lending some support to Musk's sense of urgency even if the exact dates are off.
The debate is not academic. If Musk's chart is right, the next few years will see capabilities that make today's models look primitive. If Hinton is right, the industry has years of steady improvement ahead. Either way, the conversation has moved from 'if' to 'when.'
The accuracy of Musk's chart trajectory through the rest of 2026 remains unverified.
