I'm so annoyed.

The speaker details their extensive struggles with technology, including 11 hours spent setting up a podcast and constant computer reboots. They then introduce a new "rap" song about open source and coding, which they find hilarious and a unique, self-aware take on tech culture. A significant portion of the discussion revolves around the speaker's frustration with Anthropic's Claude Code. They recount a personal experience where Claude Code failed to help debug a Dropbox issue on a new machine, despite the problem being within its supposed capabilities. The AI initially claimed the issue was outside its area of expertise, only offering to do "software engineering research." This contrasted sharply with CodeX, which successfully diagnosed and provided steps to fix the Dropbox problem. The speaker attributes this shift in Claude Code's behavior to a suspected change in its system prompt, potentially to restrict its functionality to purely software development tasks and prevent its use with third-party tools like OpenClaw. They present this as a "conspiracy theory," acknowledging the lack of concrete proof, but emphasizing the sudden and unprecedented decline in Claude Code's general utility. This is further supported by the fact that Anthropic's billing now differentiates based on whether "OpenClaw" is mentioned in the system prompt, even for paid "extra usage." The speaker concludes that Claude Code is no longer usable for their needs and announces their switch to CodeX, praising its open-source nature, transparency, and superior performance. They also mention that they believe they could win in court if Anthropic were to ban their use of T3 Code with Claude. The speaker also addresses the challenges faced by new developers in the current tech landscape. They explain that the previous era, where a high demand for developers allowed even below-average individuals to secure jobs and learn on the payroll, is over. Now, with an abundance of graduates and fewer entry-level positions, new developers must demonstrate above-average competence and likability to stand out. They introduce a concept of "average" developer skill, noting that many people's perception of average is skewed by their immediate surroundings. They illustrate this with an anecdote about their own early career at Twitch, where they were hired on a contract despite bombing technical interviews, largely due to high urgency for hires and a good "vibe match" with the team. This kind of opportunity, they argue, is now rare. The key advice for new developers is to focus on two controllable factors: competence and likability. To improve competence, developers must actively seek out communities and environments where they are surrounded by more skilled individuals. This helps in accurately measuring one's own abilities and avoiding the "Dunning-Krueger effect" (overestimating one's skill) or imposter syndrome (underestimating one's skill). The speaker emphasizes that this "surrounding" can be virtual, citing their own community as an example. Regarding likability, the speaker encourages genuine engagement and appreciation for others' work. Instead of spamming portfolios or asking for reviews, new developers should show genuine interest, ask thoughtful questions, and express gratitude to creators of tools and projects they admire. They share personal anecdotes of how this approach led to valuable connections and career opportunities, including the story of Basim, a developer who built a two-star open-source library and was subsequently referred to a Y Combinator startup, landing a full-time engineering role after a single interview. This, they argue, is because Basim demonstrated competence and genuine interest, leading to a referral based on observed skill rather than a direct request for a job. The speaker stresses that in an increasingly AI-driven and "sloppy" tech world, the human touch, genuine interest, and gratitude are more valuable than ever. They conclude by urging new developers not to lose hope, to follow their excitement, and to be good humans in the community.