{"id":542881,"date":"2026-07-10T15:10:10","date_gmt":"2026-07-10T15:10:10","guid":{"rendered":"https:\/\/www.newsbeep.com\/ie\/542881\/"},"modified":"2026-07-10T15:10:10","modified_gmt":"2026-07-10T15:10:10","slug":"anthropic-wants-you-to-use-ai-to-decide-whether-or-not-you-should-use-ai","status":"publish","type":"post","link":"https:\/\/www.newsbeep.com\/ie\/542881\/","title":{"rendered":"Anthropic wants you to use AI to decide whether or not you should use AI."},"content":{"rendered":"

On Thursday, Anthropic released reflection<\/a>, a new feature that lets users track and review their Claude activity.\u00a0Available now in beta for Free, Pro, and Max users with memory turned on, reflection was designed to help users \u201creflect on how you use Claude,\u201d per Anthropic.<\/p>\n

But engineering leaders are skeptical about the feature\u2019s ability to provide truly meaningful insights, and some worry that what\u2019s meant to help users approach AI with more discernment may end up making them even more reliant on AI suggestions and advice.\u00a0<\/p>\n

As Sergey Matikaynen<\/a>, co-founder and CTO of GoGloby, tells The New Stack: \u201cA tool that makes you think about your usage doesn\u2019t improve your judgment.\u201d<\/p>\n

Reflection tracks usage, not what developers really want<\/p>\n

In the reflection dashboard, Claude users can review their activity over the past 1, 3, 6, or 12 months. A high-level summary outlines what they use Claude for most, including key topics and usage patterns.<\/p>\n

The feature even highlights specific examples of how people use Claude, \u201clike noting that you often rework email drafts in your own voice, or delegate tasks only after settling the strategy yourself,\u201d Anthropic explains.<\/p>\n

But what the AI company presents as a catalyst to \u201creflect on and refine how you use Claude,\u201d some say, fails to deliver enough depth to effect real analysis or improvement.\u00a0<\/p>\n

\u201cA tool that makes you think about your usage doesn\u2019t improve your judgment.\u201d<\/p>\n

For example, when asked whether reflection-style usage dashboards could help developers make better decisions about when to use AI, Anish Agarwal<\/a>, co-founder and CEO of Traversal and assistant professor at Columbia University, tells The New Stack he sees the potential but doesn\u2019t believe Anthropic\u2019s version captures the right signals:\u00a0<\/p>\n

\u201cKnowing I used AI for 80% of my coding doesn\u2019t change how I work.\u201d<\/p>\n

What he wants, instead, is a system to connect coding behavior to production outcomes to help answer questions like, \u201cWhich AI-generated changes led to incidents? Which ones required significant human review? Where did engineers spend hours debugging code they accepted in seconds?\u201d It\u2019s a growing problem for new manager-like engineers who need better signals to evaluate AI\u2019s impact on engineering outcomes<\/a>.\u00a0<\/p>\n

Matikaynen shares Agarwal\u2019s interest in AI-usage insights but agrees that reflection itself is too high-level to be useful. He calls for more outcome-focused insights that capture signals like how long it takes to review AI-assisted pull requests.\u00a0<\/p>\n

\u201cIf they are delivery signal-based instead of engagement signal-based,\u201d he says, then tracking metrics would be more likely to help developers and other users determine whether to opt for AI or human work. But what Anthropic\u2019s offering up, at least for now, isn\u2019t enough, in his eyes:\u00a0<\/p>\n

\u201cIt\u2019s a count of activity that has nothing to do with that activity producing anything worth shipping.\u201d<\/p>\n

Deciding when to use AI requires context AI doesn\u2019t have\u00a0<\/p>\n

Anthropic says it designed reflection after interviews with Claude users<\/a> revealed a desire to better understand how to use AI. Specifically, the AI company says it hopes the new feature will help users answer questions like \u201cHow often should someone use AI?\u201d and \u201cWhen is AI suited to a task, and when is it better left to a human?\u201d<\/p>\n

To that end, reflection does more than summarize usage; it also provides suggestions on how to use Claude more effectively, \u201clike starting a Project instead of needing to re-explain the context of ongoing work,\u201d says Anthropic.\u00a0<\/p>\n

But engineering leaders say answering these questions and determining when and how to delegate work to AI shouldn\u2019t be left up to any vendor or system.\u00a0<\/p>\n

When asked where that guidance should come from,\u00a0Chris Wysopal, co-founder and chief security evangelist at Veracode, tells The New Stack: \u201cAI vendors should explain their models\u2019 strengths and limitations, but they shouldn\u2019t be the final authority on when to use AI.\u201d That decision, he says, should be informed by an organization\u2019s own engineering and security leaders and enforced by an independent governance layer.<\/p>\n

Moreover, Agarwal points out that AI tools like Claude are, in fact, unable to help users make decisions about AI usage as they lack the greater context and organizational knowledge to make nuanced evaluations:\u00a0<\/p>\n

\u201cThe model only knows the conversation you\u2019re having with it,\u201d he explains \u2014\u00a0a drop in the bucket compared to an engineering organization\u2019s years of production history, architectural knowledge, and incident data. \u201cThose are the signals that should determine where AI is appropriate and where human judgment is still essential,\u201d he says.<\/p>\n

The risk of using AI to judge AI usage<\/p>\n

And then there\u2019s the irony of letting an AI tool prompt users to consider whether they should use AI for a task or rely on human thinking.<\/p>\n

\u201cTheir [Anthropic\u2019s] business model depends on usage increasing, and guidance on restraint from the party who profits from increased usage is always going to be softer than the situation calls for,\u201d points out Sergey.<\/p>\n

Note that the reflection dashboard itself doesn\u2019t explicitly tell users whether they should or shouldn\u2019t use Claude. Rather, as Anthropic describes it, the feature \u201cinvites you to step back and examine the role Claude plays\u201d by occasionally surfacing questions like \u201cWhat\u2019s one thing you want to keep doing yourself, even if Claude could do it faster?\u201d and offering the chance to talk through the decision with Claude.\u00a0<\/p>\n

But Agarwal agrees that Anthropic has a clear incentive to keep users engaged with Claude, which begs the question: Is talking to an AI tool about AI usage even helpful?\u00a0<\/p>\n

More worryingly, he cautions about the risk that users will eventually optimize for engagement rather than engineering quality. If the reflection dashboard becomes just another set of numbers to monitor and potentially optimize, it could end up steering users even further away from real engineering goals.<\/p>\n

\t<\/p>\n

\n\t\t\t\tYOUTUBE.COM\/THENEWSTACK\n\t\t\t<\/p>\n

\n\t\t\t\tTech moves fast, don’t miss an episode. Subscribe to our YouTube
\n\t\t\t\tchannel to stream all our podcasts, interviews, demos, and more.\n\t\t\t<\/p>\n

\t\t\t\tSUBSCRIBE<\/p>\n

\t<\/a><\/p>\n

Group
\n Created with Sketch.<\/p>\n

\t\t<\/p>\n

\t\t\t\t\t<\/p>\n

\n\t\t\t\t\t\t\tMeredith Shubel is a technical writer covering cloud infrastructure and enterprise software. She has contributed to The New Stack since 2022, profiling startups and exploring how organizations adopt emerging technologies. Beyond The New Stack, she ghostwrites white papers, executive bylines,…\t\t\t\t\t\t<\/p>\n

\t\t\t\t\t\tRead more from Meredith Shubel\t\t\t\t\t\t<\/p>\n

\t\t<\/a><\/p>\n","protected":false},"excerpt":{"rendered":"On Thursday, Anthropic released reflection, a new feature that lets users track and review their Claude activity.\u00a0Available now…\n","protected":false},"author":2,"featured_media":542882,"comment_status":"","ping_status":"","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[6],"tags":[61,60,80],"class_list":["post-542881","post","type-post","status-publish","format-standard","has-post-thumbnail","category-technology","tag-ie","tag-ireland","tag-technology"],"_links":{"self":[{"href":"https:\/\/www.newsbeep.com\/ie\/wp-json\/wp\/v2\/posts\/542881","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.newsbeep.com\/ie\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.newsbeep.com\/ie\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.newsbeep.com\/ie\/wp-json\/wp\/v2\/users\/2"}],"replies":[{"embeddable":true,"href":"https:\/\/www.newsbeep.com\/ie\/wp-json\/wp\/v2\/comments?post=542881"}],"version-history":[{"count":0,"href":"https:\/\/www.newsbeep.com\/ie\/wp-json\/wp\/v2\/posts\/542881\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.newsbeep.com\/ie\/wp-json\/wp\/v2\/media\/542882"}],"wp:attachment":[{"href":"https:\/\/www.newsbeep.com\/ie\/wp-json\/wp\/v2\/media?parent=542881"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.newsbeep.com\/ie\/wp-json\/wp\/v2\/categories?post=542881"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.newsbeep.com\/ie\/wp-json\/wp\/v2\/tags?post=542881"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}