I would pick Meta. It already has a huge advertising business where AI can translate relatively quickly into better recommendations, targeting and ad conversion. That gives Meta a clearer path from AI investment → engagement/ad performance → revenue. The real question is whether those gains can justify its enormous capex bill.

Tesla may ultimately have the larger upside if autonomy and robotaxis work at scale, but the payoff is more dependent on future execution while margins and cash flow remain under pressure.

SpaceX is particularly interesting because Starlink can help finance its rapidly expanding AI ambitions, but its AI infrastructure spending is also extremely heavy.

So for me: Meta is most likely to prove measurable AI returns first, Tesla has the biggest “if autonomy works” optionality, and SpaceX is the wildcard.