It's only a slight oversimplification to say that the global economy is now a wrestling match between an energy shock and a tech boom. On one side is the oil shock radiating from the US war on Iran, which has pushed Brent crude back above $90 a barrel and revived concerns of a return to April's $120 highs. On the other is a spectacular, but unevenly distributed, AI boom.

RBI's MPC, meeting in the first week of August, will set its policy rate chiefly on domestic considerations, as it should. But external conditions often matter on the margin. So, it's worth reviewing the dilemma they pose for the economy.

Start with the energy shock. The war has entered its fifth month with an escalation, not a resolution. A ceasefire announced weeks ago is in tatters. Strikes and counter-strikes have again sealed the Strait of Hormuz, while a disruption of Red Sea shipping by Houthi forces could well be on the cards. Global inventories, drawn down sharply since the war began, leave far less cushion than at the outset, arguably making this an even more perilous inflection point than April.

Exposure to the shock is uneven. The US, somewhat ironically, is one of the least exposed countries because it is a net energy exporter, although it also has lower petroleum taxes than other advanced economies and, so, higher percentage pass-through.

Europe and emerging Asia, more reliant on natural gas - a market with fewer substitutes when a specific pipeline is disrupted - have seen retail prices climb further. In China, prices have been somewhat cushioned by large inventory drawdowns and a more substantial shift towards renewables.

The energy shock explains why global growth has come under pressure this year, and the tech boom explains why it has not completely buckled under. The boom is centred in the US, where the astonishing bull run of Nasdaq and S&P has relied heavily on AI-related companies, and where business investment related to AI has been a crucial prop for domestic activity. When taken together with greater energy independence, it's not surprising that the US remains the major growth engine among advanced economies.

But the AI boom extends much further, to countries such as South Korea, Japan and Taiwan. Even in China, where the lingering impact of a property bust and inadequate domestic demand have resulted in Q2 growth of 4.3% - lowest in 5 yrs - the saviour has been a surge in AI-related exports.

Unfortunately, risks to this state of play are asymmetric, pointing in the direction of higher inflation and lower growth. Inflation expectations have stayed largely anchored so far, with markets treating the oil shock as transitory. But that composure is being tested by events. Sovereign yields have risen broadly.

In June, European Central Bank raised rates for the first time in 3 yrs, and Bank of Japan lifted its policy rate to 1%, the highest level since 1995. In the US, Fed governor Christopher Waller has warned that another hot core-inflation reading - core CPI is running at 3.1% - could force a hike the Fed had not planned to deliver.

On the growth front, a tech boom - this narrow - is precarious, for two reasons:

• There's justifiable scepticism about whether market valuations of everything AI-related are sustainable. Both Nasdaq and S&P have been flat or declined slightly since the beginning of June, perhaps reflecting the political backlash against energy-hungry data centres. Moreover, investors are increasingly concerned about the impact of rising policy rates on AI valuations.
• This fear is almost the opposite of the first: the AI boom will cannibalise businesses it does not feed. IBM provides the clearest illustration yet. Its shares fell by 25% in a single mid-July session, the worst day in the company's 115-yr history. This was the result, its CEO admitted, of clients redirecting budgets towards AI infrastructure and away from the services that IBM has long sold. Either a market correction or a widening squeeze on non-AI business could further weaken global economic activity.

Implications for India are anything but straightforward. Sputtering external demand would strengthen the case for a rate cut, and higher prices the case for a rate hike. As with most stagflationary impulses, there is no clean textbook reaction here.

Monetary policy should usually aim to stay ahead of the curve, considering lags in the transmission of policy rates to real economic activity. But in this case, the most instructive lesson from the global picture may be how difficult it is to plausibly extrapolate. Oil prices could retract to $60, or soar past $120. The tech boom could correct sharply, or continue supporting growth for another year. The uncertainty is almost Knightian.

On balance, with risks to inflation pulling in the opposite direction from the risks to growth, external conditions are most consistent with a wait-and-see attitude. RBI should act as and when external shocks push the domestic data unambiguously in a certain direction, not until then.