Mostly Technical
150: Everybody Loves Jev
· Summarized
Key takeaways
- Understanding human motivations and trade-offs is essential for effective sales and marketing, especially in technical fields.
- Most SaaS products will continue to be used by large companies, even as some categories may experience erosion due to AI automation.
- Jev is a promising AI classification tool that is cheaper and faster than general-purpose LLMs, with potential uses in workflow automation and decision-making.
- Responsible use of AI, including managing reliability and accuracy, is critical, especially when integrating new models like Jev into production.
- Timing for raising prices should consider inflation and market conditions; many companies could and should raise prices, but often do not.
- Pricing decisions depend heavily on customer value, reliability, and perceived quality, not just on cost or competitors’ prices.
- In enterprise-level software, long-term trust and deliverability often outweigh short-term cost savings.
- Despite debates on industry shifts, the idea that entire SaaS sectors are vanishing is overblown; most users still rely on established tools.
Summary
Solo Parenting and Tech Reflections
Ian starts the episode describing his weekend as a solo parent to four kids under six, sharing humorous anecdotes like forgetting to put shorts on his daughter and how the older kids are starting to help more. He reflects on the logistics of solo parenting getting easier, especially as his children help with tasks like getting ready for church or grabbing items. Despite the exhaustion, he finds the experience rewarding and appreciates the moments alone with his children, emphasizing the value of occasional solo time. He also discusses his weekend activities, including editing photos of his brother's horse show using Jev, a new AI classification model, which he finds invaluable. Ian humorously criticizes the overhyping of AI tools on Twitter, expressing frustration with people misunderstanding economic fundamentals and over-exaggerating AI's impact on industries. There’s a lively debate about the real-world effects of AI on SaaS companies and whether companies are starting to build or replace tools in-house using AI, with skeptical views on the industrywide shift.
Ian's Twitter Frustration and Microeconomics Rant
Ian passionately criticizes the misunderstandings prevalent on Twitter about microeconomics, demand-supply, and trade-offs, asserting that many people are making poor economic decisions due to a lack of understanding. He jokes about the unlikelihood of anyone creating a comprehensive microeconomics course from social media content and emphasizes the importance of understanding human motivations for effective sales strategies. The discussion intensifies over tweets by Jon Yongfook regarding SaaS moats and automation, with Ian arguing against the idea that entire SaaS categories are disappearing. They analyze specific cases like Mintlify, BannerBear, and company decisions to build or buy tools, concluding most businesses will stick with existing solutions due to efficiency and risk factors. The conversation underlines that while some industries may change, the idea of SaaS vanishing industry-wide is overblown, especially for enterprise-level, high-value software.
Jev — A New Classification Model for AI Products
The episode introduces Jev, a new AI system designed for classification tasks, which only produces structured responses instead of free-form text, making it cheaper and faster than traditional LLMs like Codex or GPT. Ian and Aaron discuss how Jev can be used to replace heuristics, improve workflow automation, and enable cost-effective decision-making in product development. They explore practical uses of Jev, such as for spam detection, email filtering, and classification of user inputs, arguing that it could handle large volumes at a fraction of the cost of conventional models. Both agree that while Jev is promising, it’s still in early days, with concerns about its availability and reliability. They consider how product builders could leverage Jev as an efficient heuristic, how it might be integrated into existing workflows, and its potential limitations in accuracy. The segment emphasizes responsible, thoughtful use of AI classification tools, balancing speed, cost, and correctness.
Pricing Strategies in a Changing Market
Aaron and Ian debate whether now is a good time for software companies to raise or lower prices amid inflation and AI-driven efficiency gains. Ian advocates for raising prices, citing industry examples like QuickBooks and Salesforce, stressing that most businesses could easily increase prices by 20% and still retain customers due to existing competitive gaps. They also discuss the psychological and practical aspects; Ian advises that smaller bootstrap companies should consider raising prices to match inflation, but acknowledges that many miss this opportunity. Jesse Hanley’s position on not raising prices is contrasted, with the debate revealing that the decision to charge more depends on the customer base—developers might be more price-sensitive, whereas larger enterprises value reliability and deliverability. The conversation concludes that price increases are justified by improved product value, especially with AI integration, but cautions that many small companies neglect to adapt their pricing strategies over time. Ultimately, both agree that strategic pricing, informed by understanding customer motivation and industry dynamics, is crucial for sustainable growth.
Sponsors
- Svix · Svix has solved all the hard problems about scaling web hooks so you can avoid triaging failed deliveries and incrementally building features. They are the official level job board, and more tech details are in the show notes.
- Bento · Bento is the email marketing and transactional email platform built for the AI era, offering rates as low as nine cents per thousand emails. They also provide CRM features, perfect for product marketing and transactional outreach.
- PostShiba · PostShiba is a transactional email platform that powers Bento's delivery success. They offer raw sending powers and allow you to build your own email services, like email resets, receipts, and more, with a new developer-friendly platform.
- Valorin Security · Valorin Security emphasizes that AI helps attackers build tools and find vulnerabilities faster, so they recommend having a friendly Australian hacker, like Stephen Reese Carter, on your team to catch security holes before they are exploited.
- SerpAPI · SerpAPI offers access to Google search, maps, shopping, and other search engines via API, with data in markdown and JSON. It abstracts away the headaches of scraping, making it handy for AI tools, and it's easy to try for free.
- Typesense · Typesense provides lightning-fast, open-source search for your site or app, integrated with Laravel Scout, with no extra effort required.
Show notes
Ian and Aaron discuss the arrival of Jev, whether or not people should take microeconomics classes, Aaron's new office, raising prices, and so much more.
Sponsored by Svix, Bento, PostShiba, Valorin Security, SerpAPI, Laracon AU, and Typesense.
Interested in sponsoring Mostly Technical? Head to https://mostlytechnical.com/sponsor to learn more.
- (00:00) - Solo Parenting
- (09:00) - Ian's Very Mad About People on Twitter
- (17:12) - People Should Take Microeconomics (?)
- (31:04) - Aaron Is Leasing An Office!
- (45:22) - Jev!
- (01:04:56) - Should you charge more or less?
Links:
Follow this show on AudioBrevity and new episodes get summarized for you. Create a free account