A Qualitative Leap in AI According to Matt Shumer

<p>Today, Silicon Valley is actively discussing Matt Shumer's essay "Something Big Is Happening" (originally published in Fortune).</p>
<p>Shumer, the founder of HyperWrite, writes that the latest models (GPT-5.3, Claude Opus 4.6) are not just another incremental improvement, but a qualitative leap. He describes how it now works: "I describe what I want to build in plain English, and it just... appears" — without the need to correct the AI.</p>
<p><strong>Key points:</strong></p>
<ul>
<li>AI is now involved in creating the next versions of itself. OpenAI's documentation states that GPT-5.3 "was instrumental in creating itself".</li>
<li>According to METR, AI capabilities are doubling every 4-7 months.</li>
<li>The CEO of Anthropic predicts that 50% of entry-level white-collar jobs may disappear in 1-5 years.</li>
<li>Unlike previous waves of automation, AI is improving simultaneously across all cognitive areas — there is no "window for retraining".</li>
</ul>
<p><strong>His recommendations:</strong></p>
<ul>
<li>Implement AI in your work right now — become the one who demonstrates its value in the organization.</li>
<li>Create a financial cushion and reduce debts.</li>
<li>Set aside an hour a day for experimentation.</li>
<li>Focus on skills that require personal trust, physical presence, or licenses.</li>
</ul>
<p>What do you think?</p>
<p>Link: <a href="https://shumer.dev/something-big-is-happening">shumer.dev/something-big-is-happening</a></p>;
<p>#ai #future #work #matt_shumer</p>
<p>—————————<br>Мысли Рвачева<br>—————————</p>