Former Google CEO Eric Schmidt says: Engineers who are still coding the old way should now ...

Eric Schmidt said AI has triggered a major shift in software development since about October last year, arguing that traditional line-by-line coding is obsolete. He expects productivity to surge, enabling individuals to build powerful applications that previously required large teams, while warning investors that some software products may become easier to replace. The message is bullish for AI adoption and software efficiency, but the piece is commentary rather than a direct company-specific catalyst.

Analysis

This is less about “AI is helping programmers” and more about a near-term reset in software cost curves. If agentic coding meaningfully cuts implementation time, the first beneficiaries are the platforms that own model distribution and workflow embedding, while the first casualties are point-solution vendors whose moat is mostly feature depth and headcount-heavy services. That should widen dispersion inside software: broad index exposure is less attractive than picking firms with proprietary data, strong usage-based pricing, and products that can become the default IDE/workflow layer.

The second-order effect is margin compression for the lower end of the software stack before revenue shows up anywhere else. Buyers will try to renegotiate contracts on the assumption that replacement costs are falling, which can hit renewal rates over the next 2-4 quarters even if usage remains stable. At the same time, engineering productivity gains are bullish for infra demand: cheaper code creation tends to increase the amount of software shipped, which lifts spend on cloud, observability, security, and compute-intensive AI tooling rather than reducing it.

For GOOGL, the market is still underpricing the optionality of becoming the default productivity layer for developers, not just a model vendor. The risk is that this narrative benefits multiple incumbents at once and the moat comes down to distribution, not technology — meaning upside may be real but not exclusive. The bigger reversal trigger is model commoditization or an enterprise backlash if AI-generated code creates security/quality incidents; those failures would likely show up first in the next 6-12 months through elevated review costs and slower deployment velocity.

Contrarian view: the crowd may be extrapolating “fewer engineers” too literally, when the more probable outcome is that AI increases total software output and expands TAM. That makes blanket shorts on software quality names dangerous; the better short is on vendors with thin differentiation and labor-arbitrage services exposure. The strongest risk/reward is in owning the picks-and-shovels of accelerated software creation while fading companies whose only defense is manual implementation friction.

Key Decisions for Investors

• Long GOOGL on a 3-6 month horizon as a core AI workflow beneficiary; target a 15-20% upside re-rating if developer productivity tools become a meaningful distribution wedge, with risk capped if enterprise adoption proves incremental rather than platform-shifting.
• Pair trade: long GOOGL / short a basket of low-differentiation application software names with heavy services exposure and weak retention; expect 2-4 quarters for renewal pressure to surface if buyers use AI efficiency as a pricing lever.
• Long cloud/security infrastructure basket (e.g., MSFT, AMZN, CRWD, DDOG) vs short low-multiple legacy dev-tools where code-generation commoditization is most directly substitutive; hold for 6-12 months as productivity gains increase total software throughput.
• Consider buying GOOGL call spreads 6-9 months out to express upside from AI developer adoption while limiting premium outlay; structure for a 2:1 or better payoff if the market starts pricing a broader platform moat.
• Avoid outright shorts on broad software indices until there is evidence of AI-driven quality failures or budget deflation; the more likely near-term effect is mix shift and vendor churn, not absolute demand destruction.