Resolution increases make them more expensive, then efficiency gains reduce costs - a sawtooth pattern. But in every case, this means generating more tokens.
tomtunguz.com/hidden-cost-...
Will smarter models be increasingly expensive because of greater accuracy or less expensive because they’re smarter?
“For text, I’m seeing 1.46x more tokens for the same content. We can expect it to be around 40% more expensive in practice.” - Simon Willison
Boris Cherny, creator of Claude Code, acknowledged Anthropic raised rate limits “to make up for it.”
Smaller pieces force the model to pay closer attention to each word, like reading a contract word by word instead of skimming paragraphs. The model follows instructions more precisely & makes fewer mistakes on coding tasks. The tradeoff : more tokens, higher costs.
Then Opus 4.7 shipped & the smarter model became much more expensive. The cause : a new tokenizer - software to break text into pieces a computer understands.
The trend across vendors has been smarter models using fewer tokens per task.
When Anthropic launched Opus 4.5 in November 2025, the bigger, more expensive model was actually cheaper to use.
On a per-token basis, Opus 4.5 costs 67% more than Sonnet. But Opus 4.5 used 76% fewer tokens to reach the same outcome. A task that cost $1 on Sonnet cost $0.40 on Opus.
“As models get smarter, they can solve problems in fewer steps : less backtracking, less redundant exploration, less verbose reasoning. Claude Opus 4.5 uses dramatically fewer tokens than its predecessors to reach similar or better outcomes.”
Apple Podcasts : podcasts.apple.com/us/podcast/o...
Spotify : open.spotify.com/episode/0yET...
YouTube : www.youtube.com/watch?v=8DLp...
tomtunguz.com/lena-waters-...
We used to sell to humans who researched with tools. Now we sell to tools that report to humans. Equip your agents like you’d equip an internal champion.
In the enterprise, with many stakeholders & complexities, the path isn’t so clear. The wrong decision still falls on the person. You can’t sue an agent.
“You can’t go to your board and say, my agent told me we should do this.”
Agents are joining buying committees. For small value purchases, they are also the decision-maker. Which database should I use for my vibe-coded app? Let the agent decide. Pair of shoes? New laptop? New car? All of those decisions could be made entirely through AI.
Some companies have abandoned websites & mobile apps entirely. The replacement isn’t a better website. It’s just a wall of text in the favorite format of an agent : markdown.
“Think about how much time we’ve spent debating the top nav. Solutions before products? Are we allowed to call ourselves a platform? That’s applied human psychology. Agents don’t care.”
Websites are human artifacts. They exist to communicate, persuade & convert people who navigate to them. AI agents don’t care for beautiful styling or appeals to emotion. This new persona doesn’t browse. It parses.
Marketing teams can finally build their own tools. That unlocks growth. But it’s still phase one.
The first phase of AI transformation is debt repayment. Most companies are agentically connecting go-to-market processes that should have been fixed years ago.
“Removing human coordination overhead and calling it transformation? That’s debt repayment. It’s real value, but it’s not a new paradigm.”
No one comes into a sales conversation without first asking an AI. The buyer journey has changed. Lena Waters, marketing leader behind DocuSign’s IPO, Grammarly & Notion, joined me on Office Hours to discuss what this means for your go-to-market.
Every role fits somewhere on this 2x2. I would put venture capitalist in finite demand & open loop. There’s only a certain amount of venture capital dollars entering the ecosystem in a year, & investment selection remains an open problem.
Where does yours fit?
tomtunguz.com/ai-problem-m...
Open Loop + Finite Demand = Utility Tools. Preparing 10-Ks & 10-Qs. Legal contract review. Insurance claims processing. One report per quarter, one contract per deal. AI makes the work faster, but doesn’t create new work to do.
A person must judge the right ones to publish. Does this ad campaign align with our values? Is this strategic positioning correct? Some problems are open loop today but will close over time.
Open Loop + Infinite Demand = Creative Amplifiers. Content creation & marketing strategy. AI can generate a thousand ad variations or blog posts.
Closed Loop + Finite Demand = Efficiency Plays. AI bookkeeping categorizes transactions, reconciles accounts, files returns. Deterministic rules applied to numbers. But a company only has so many transactions. A company files taxes once a year. It closes the books each quarter.
Closed Loop + Infinite Demand = Economic Engines. Software engineering lives here. AI writes the code. Tests verify correctness. More code enables more features. Companies will always need more software.
On the other axis, open vs closed loops. Closed Loop means AI can verify correctness without human intervention.
But that’s not true for all roles. I use a 2x2 matrix that separates work along two axes : the ceiling of demand & whether the loop can be closed.
On one axis, demand. Infinite Demand means more output creates more value. There is no saturation point.
Now, it’s 275 million per week, on pace for 14 billion this year if growth remains linear (spoiler : it won’t.) GitHub Actions has grown from 500M minutes/week in 2023 to 1B minutes/week in 2025, and now 2.1B minutes so far this week.
The demand for software is infinite. Kyle Daigle, GitHub’s COO, made the case concrete :
There were 1 billion commits in 2025...
AI is breaking every system it touches : data centers, financial markets, security defenses. Software was lunch. What’s for dinner?
tomtunguz.com/mythos-glass...
Engineering budgets redirect. A significant fraction of AI tokens spent on software development will shift to hardening. Every company shipping code will need to scan it at this level of sophistication. Buyers will start to demand this level of hardening.