Arvind Krishna's message alongside IBM's second quarter results was that the company sits at the start of something rather than the wrong end of it.
The chief executive said IBM fundamentally believes business is in the early innings of a structural shift, and that its portfolio across software, infrastructure and consulting is positioned to help clients capture the value of an artificial intelligence driven future.
The difficulty is that the same quarter offered a fairly persuasive case for the opposite reading.
Sales of IBM's Z mainframe computers, the transaction processing systems used by banks and airlines, fell 42%, dragging infrastructure revenue down 7%.
Software growth halved to 5% from 11% in the first quarter, and consulting was flat.
If corporate customers really are redirecting budgets towards AI data centre equipment, IBM is currently on the paying side of that trade rather than the receiving side.
Finance chief James Kavanaugh's explanation is essentially a timing argument.
He said customers are building out infrastructure now and will monetise it later, and that the shock was one of magnitude rather than direction, driven by extreme price increases from hardware and memory suppliers over the past 30 days.
He also disputed the more alarming interpretation doing the rounds, that AI is beginning to eat conventional software spending, describing the shift as clients simply filling out an infrastructure layer.
That distinction matters well beyond IBM, which is why some analysts were relieved.
CFRA's Brooks Idlet said the print should be read as a positive for the broader software sector, on the basis that IBM's problems look company specific.
But a timing argument only works if the timing eventually arrives.
IBM told investors in April it expected constant currency revenue growth above 5% for the year, underpinned by software growing more than 10%, and both of those claims have now been withdrawn.
Guidance has been cut to 4% to 5%, and Krishna conceded last week that the company had faltered in adapting, with numerous large deals slipping.
A management team that misjudged its own quarter by roughly $660 million of revenue is asking to be believed about the shape of the next several years.
The interim answer is cost.
IBM will lift its savings programme above the $5.5 billion annual exit run rate previously targeted, which Kavanaugh said delivers about 300 basis points of operating leverage from productivity alone.
That is how the company grew profit 5% and expanded operating margins on 1% revenue growth, and it buys time.
It does not settle the question of whether IBM's integrated model, which Kavanaugh again defended against break-up speculation, is the right structure for the shift he is describing.
Thomas Martin of Globalt Investments put the market's position bluntly, saying IBM will sit in the penalty box for a while as it adjusts.
The roughly 2% after hours rise suggests investors were reassured that the second quarter contained no fresh disasters, not that they have accepted the early innings thesis.
On that, IBM will need several quarters of evidence, and the mainframe cycle to turn, before the benefit of the doubt returns.
