AI's threat to incumbent tech makes SpaceX the preferred investment for the next decade. AI has begun taking business from Google, with Amazon and Apple next; Microsoft's 27% workforce reduction program signals desperation. x.com 

Salesforce might be dead as intelligence becomes an abundant commodity, though it can argue it has distribution. Claude, OpenAI and open-weight models such as Kimi, Qwen and DeepSeek are in a price fight to zero, so years of developer work no longer form a moat. The new moats are trust, distribution and data, especially data behind corporate firewalls, in hospitals or otherwise hard to get. x.com 

The SaaSpocalypse narrative makes less and less sense for Salesforce, ServiceNow and Palantir, because enterprises will deploy AI agents that use software tools rather than build their own software. Jensen Huang says agents won't replace tools but will use them, and NVIDIA has deployed so many agents that its tool usage is actually increasing. x.com 

AI embedded throughout software could bring huge future token spend to Amazon and Alphabet, whose AWS and GCP clouds host software builders and offer AI platforms. Ramp data show VC-backed firms spend 10 times more per employee on AI monthly than PE-owned and other firms, a possible leading indicator of software using more tokens relative to people. VC-driven demand is vulnerable to capital-cycle swings, but longer term incumbent software may become AI-native or be replaced. x.com 

Application software makers Adobe, Salesforce and ServiceNow are sneaky losers from the AI-lab slowdown, because frontier labs can now focus on their customers instead of AGI. Lab leaders including Elon Musk, Dario Amodei and Sam Altman have pivoted to warning that AI could kill you, which, with AI already polling badly, will make it much harder to attract investment. The fallout could be real. x.com 

These software names get stronger as AI grows, despite the early-2026 "AI will kill SaaS" sell-off. ServiceNow is becoming the control room that manages companies' AI agents. Salesforce's Agentforce brings in $1.2B a year and is tripling; Atlassian clients using Rovo spend twice as much; Adobe's business-safe Firefly earns $250M; Workday's agents bring nearly $500M. SAP and Snowflake hold the data AI needs, and Datadog monitors the extra systems AI creates, with the AI wave still early. x.com  … Show more

Software stocks probably have a floor as results challenge fears that AI agents will displace them. Since the June 22 momentum unwind, IGV is up 25% and the SOX index down 22%. Atlassian and Salesforce jumped on solid results, plus an Anthropic deal for Salesforce, and possible Silver Lake interest lifted Workday. A new bullish twist says agents will use software tools 10-100x more than humans. Whether fundamentals back the rally is still open; a likely Fed hike and weak September are risks. … Show more

AI-first software like Palantir, Zeta Global and Duolingo is still largely misunderstood and underappreciated, while data-light SaaS is dying by a thousand cuts. Data-light SaaS may keep large enterprises on convenience and switching costs, but will struggle to win new customers from far cheaper alternatives with 90% of the capability, capping its growth. Per The Information, small companies are replacing Salesforce and HubSpot with custom AI-built apps, cutting software costs 40-80%. x.com 

Microsoft, ServiceNow and Salesforce could capture the next AI wave because it will be software, not chips. Nvidia's CEO said there will be 100x more agents using enterprise tools, making software vendors the real bottleneck. After 2025's winners Nebius, SanDisk and CoreWeave, the next trillion dollars may go to companies charging AI agents monthly subscriptions. x.com 

Salesforce is exposed to AI-native rivals aiming to own the sale itself rather than build a better CRM. Upstarts such as Attio, Monaco and Rox can capture the decision traces behind closed deals, a learning loop Claude cannot rebuild from Salesforce's fields. Its schema is more tax than foundation: relationship-driven work like recruiting and VC does not fit it, and young startups cannot be served profitably. Its 150k customers won't churn soon, but new AI-native firms may never install it. … Show more