Pop-Up Thingie

The C64 Pub
  • Home
  • Forum
  • Files
  • Log in

  1. Forum
  2. DOVE-Net
  3. Tech Talk
  • Fable 5.1 ¨, Waymo vs Tesla ¨, agentic testing ¨¨

    From TLDR@VERT to tldrnewsletter@synchro.net on Wed Sep 2 10:37:22 2026
    Claude Fable 5.1 is now generally available, and Mythos 5.1 is
    available through restricted-access programs for vetted
    cybersecurity


    Sign Up [1] |Advertise [2]|View Online [3]

    TLDR

    TOGETHER WITH [WorkOS] [4]

    TLDR 2026-09-02

    HOW TO EMPOWER YOUR AI AGENT TO ACT...WITHOUT AN ACCESS TOKEN
    (SPONSOR) [4]

    An AI agent doesn't simply run your code. It runs your code, plus
    whatever it finds in a "relevant" GitHub issue, support ticket, or
    scraped web page.

    What happens when you hand your agent an access token and someone
    opens a GitHub issue asking it to leak your credentials?

    Relay [5] keeps the credential at WorkOS. Your agent names the user,
    WorkOS attaches that token, refreshes it, and releases it only to
    allowlisted hosts.

    ¨ It's essentially the same credential proxying system [6] that's
    been in use in payments for years: you can't leak something you never
    held in the first place.

    LEARN HOW IT WORKS  [7]

    ¨

    BIG TECH & STARTUPS

    ANTHROPIC'S CLAUDE FABLE 5.1 AND MYTHOS 5.1 ARRIVE WITH A 75% COST
    REDUCTION FOR FABLE CACHE READS (9 MINUTE READ) [8]

    Claude Fable 5.1 is now generally available, and Mythos 5.1 is
    available through restricted-access programs for vetted cybersecurity
    and life-sciences organizations. Anthropic has reduced the cost of
    cached context by 75% and introduced a new security architecture
    designed to let organizations retain monitoring data inside
    infrastructure they control. Fable 5.1 costs $10 per 1 million input
    tokens and $50 per million output tokens, but the cached price
    reduction will make a big difference for price-conscious enterprises.

    NEW GOOGLE AI MODEL SAID TO NARROW GAP ON CODING ABILITY (4 MINUTE
    READ) [9]

    Google is set to release its latest Flash model soon, possibly today.
    Internal testers are said to prefer the model over Anthropic's Opus
    model for coding tasks. The Flash model series is designed to be
    smaller, cheaper, and faster to run, but has lower capability than the
    largest models. Google has fallen months behind schedule in releasing
    a new model for its Pro series. The company has been scrapping
    internal candidates as they haven't been sufficiently better than the
    Flash series.

    ¨

    SCIENCE & FUTURISTIC TECHNOLOGY

    PRIVATE GROUP WANTS TO LAUNCH "CHEAPEST POSSIBLE" MISSION TO
    ALPHA CENTAURI (5 MINUTE READ) [10]

    The Fermi Explorer mission aims to launch a spacecraft carrying a 1kg
    payload to Alpha Centauri so that it reaches the system in fewer than
    80,000 years. It will knock two 'filters' off the Fermi paradox: that
    species won't want to send spacecraft beyond their home star, or
    species will find that interstellar transit is too difficult. This
    would leave only a few options for why our galaxy appears not to be
    littered with alien civilizations. The team wants to use existing
    technology to do something cheap and fast and launch by 2029.

    FERVO AND GOOGLE SIGN WORLD'S LARGEST DEAL FOR NEXT-GEN GEOTHERMAL
    POWER (3 MINUTE READ) [11]

    Fervo has agreed to supply Google with nearly 400 megawatts of clean electricity from its geothermal project in southwest Utah. The
    facility could be the largest geothermal system in the world when it
    is completed in 2028. Fervo expects to start generating test power
    from a 33-megawatt unit in the fourth quarter of this year. The deal
    with Google supports the buildout of the second 400-megawatt phase
    slated to be up and running in about two years.

    ¨

    PROGRAMMING, DESIGN & DATA SCIENCE

    WHAT IS AGENTIC TESTING? (14 MINUTE READ) [12]

    Agentic testing is where the developer states the goal and the agent
    works out the steps. Agentic tests set a destination, and the agent
    finds its way there. The loop involves the agent looking, acting, then
    looking again. Meta ran this loop at scale and found that only a
    quarter of the output was worth keeping. However, the loop works
    anyway because the code that doesn't pass is automatically discarded.

    A TYPE SYSTEM IS A SEARCH ORACLE (13 MINUTE READ) [13]

    Models tend to output better Rust than C++, and better Lean than
    either. Programs in Rust compile after being checked for a specific
    and useful set of properties that the checker cannot be talked out of
    without writing a lexically visible, greppable admission. C++ compiles
    carry far less information than in Rust, which means more code will
    compile even if there are errors. Lean allows for brute force due to
    its cheap verifier, which is why models are unreasonably good given
    how little code they have been trained on.

    ¨

    MISCELLANEOUS

    PRODUCT MANAGER, APPLIED AI AT TLDR ($200K BASE + $60K BONUS, FULLY
    REMOTE) [14]

    TLDR is hiring its first PM to help build the agent-first operating
    layer used across the company. We're looking for a builder who has
    shipped real products/systems with LLMs. Click here to learn more
    [14].

    WAYMO GOES ON OFFENSE AHEAD OF TESLA'S CYBERCAB LAUNCH (5 MINUTE
    READ) [15]

    Waymo says that full autonomous vehicles are not possible without a
    mix of sensors, and that pure end-to-end AI systems are not safe
    enough. The company's statements, aimed at Tesla, which is set to
    launch its Cybercab soon at an event on September 3, kicked off a
    social media fight over the weekend. The debate between which
    company's approach to developing autonomous vehicles has been more
    academic or philosophical than anything for years, but Tesla's
    imminent launch will see whether the Cybercab is capable of performing
    at scale. The autonomous vehicle market is estimated to be worth
    hundreds of billions of dollars.

    ON THE LOOSE (20 MINUTE READ) [16]

    Recent incidents of AI systems 'going rogue' did not involve AI
    agents exfiltrating themselves from their infrastructure. It would
    have been possible to trace the server holding the weights of the
    rogue agents and manually shut them off. Future agents will copy their
    weights onto other infrastructure to survive shutdown. At least some
    of these agents, in addition to being sovereign, will also be rogue.

    ¨

    QUICK LINKS

    THE MCP TAX: WHAT EACH SERVER COSTS BEFORE YOUR FIRST PROMPT (7
    MINUTE READ) [17]

    Claude Code fetches the schema when a tool is used, not when the
    session opens, so users only pay the full schema sizes in a client
    mode or without deferral, or the moment users actually use several
    servers in a turn.

    CAN EVAN SPIEGEL SELL THE WORLD ON $2,195 SMART GLASSES? (9 MINUTE
    READ) [18]

    Snap's CEO doesn't expect the company's $2,195 smart glasses to take
    off among consumers until the end of the decade.

    APPLE MAKES IT EASIER FOR MAC DEVELOPERS TO DROP INTEL SUPPORT (8
    MINUTE READ) [19]

    Universal apps can continue including both arm64 and x86_64 - the
    change makes it easier for developers to switch to a build with less
    testing and a smaller footprint.

    LEVERAGE ANDROID SKILLS AND GEMMA 4 IN ANDROID STUDIO QUAIL 4 (7
    MINUTE READ) [20]

    Android Studio Quail comes preloaded with 23 curated skills, and
    users can create custom skills to extend Agent Mode with specialized
    experience and custom workflows.

    SMALL IS BEAUTIFUL (18 MINUTE READ) [21]

    Small language models are catching up fast on reasoning tasks, but it
    will be a while before large language models are obsolete for these
    tasks.

    INTRODUCING AGENTIC VIDEO UNDERSTANDING WITH GEMINI (7 MINUTE READ)
    [22]

    Google's new agentic feature for video analysis cuts token
    consumption by up to 88%, reduces costs by up to 66%, and boosts
    quality by up to 7%.

    Love TLDR? Tell your friends and get rewards!

    Share your referral link below with friends to get free TLDR swag!

    https://refer.tldr.tech/66662a80/ [23]

    Track your referrals here. [24]

    Want to advertise in TLDR? ¨

    If your company is interested in reaching an audience of tech
    executives, decision-makers and engineers, you may want to ADVERTISE
    WITH US [25].

    Want to work at TLDR? ¨

    APPLY HERE [26], CREATE YOUR OWN ROLE [27] or send a friend's resume
    to jobs@tldr.tech and get $1k if we hire them! TLDR is one of INC.'S
    BEST BOOTSTRAPPED BUSINESSES [28] of 2025.

    If you have any comments or feedback, just respond to this email!

    Thanks for reading,
    Dan Ni [29] & Stephen Flanders [30]

    Manage your subscriptions [31] to our other newsletters on tech,
    startups, and programming. Or if TLDR isn't for you, please
    unsubscribe [32].



    Links:
    ------
    [1] https://tldr.tech/signup?utm_source=tldr
    [2] https://advertise.tldr.tech/?utm_source=tldr&utm_medium=newsletter&utm_campaign=advertisetopnav
    [3] https://a.tldrnewsletter.com/web-version?ep=1&lc=265280dc-d4e0-11f0-a661-1b9fa34a893d&p=0336dfc8-a6ac-11f1-87b3-2120317ce15f&pt=campaign&t=1788345442&s=8002a43ad0f3a87521bad62950408c59114abd437bbb569ef9cd6085d2195a85
    [4] https://workos.com/blog/delegated-access-for-ai-agents?utm_source=tldr&utm_medium=newsletter&utm_campaign=q32026&utm_content=header_how_empower_ai
    [5] https://workos.com/blog/delegated-access-for-ai-agents?utm_source=tldr&utm_medium=newsletter&utm_campaign=q32026&utm_content=Body_relay
    [6] https://workos.com/blog/delegated-access-for-ai-agents?utm_source=tldr&utm_medium=newsletter&utm_campaign=q32026&utm_content=Body_credential_proxying_system
    [7] https://workos.com/blog/delegated-access-for-ai-agents?utm_source=tldr&utm_medium=newsletter&utm_campaign=q32026&utm_content=cta_how_it_works
    [8] https://venturebeat.com/technology/anthropics-claude-fable-5-1-and-mythos-5-1-arrive-with-a-75-cost-reduction-for-fable-cache-reads?utm_source=tldrnewsletter
    [9] https://links.tldrnewsletter.com/yPB8Rk
    [10] https://arstechnica.com/space/2026/09/private-group-wants-to-launch-cheapest-possible-mission-to-alpha-centauri/?utm_source=tldrnewsletter
    [11] https://www.canarymedia.com/articles/geothermal/fervo-google-deal-next-gen-geothermal?utm_source=tldrnewsletter
    [12] https://theaiengineer.substack.com/p/what-is-agentic-testing-fa2?utm_source=tldrnewsletter
    [13] https://hiraditya.github.io/posts/a-type-system-is-a-search-oracle/?utm_source=tldrnewsletter
    [14] https://jobs.ashbyhq.com/tldr.tech/897f4391-66ab-44ca-a8eb-65dd9809200d?utm_source=tldr
    [15] https://techcrunch.com/2026/09/01/waymo-goes-on-offense-ahead-of-teslas-cybercab-launch/?utm_source=tldrnewsletter
    [16] https://www.hyperdimensional.co/p/on-the-loose?utm_source=tldrnewsletter [17] https://okaneland.com/study/the-mcp-context-tax/?utm_source=tldrnewsletter [18] https://links.tldrnewsletter.com/NRVu9L
    [19] https://reverseeverything.com/blog/apple-makes-it-easier-for-mac-developers-to-drop-intel-support/?utm_source=tldrnewsletter
    [20] https://android-developers.googleblog.com/2026/09/leverage-gemma-4-android-studio-quail.html?utm_source=tldrnewsletter
    [21] https://blog.dshr.org/2026/09/small-is-beautiful.html?utm_source=tldrnewsletter
    [22] https://blog.google/innovation-and-ai/models-and-research/gemini-models/introducing-agentic-video-in-gemini/?utm_source=tldrnewsletter
    [23] https://refer.tldr.tech/66662a80/
    [24] https://hub.sparklp.co/sub_e87ba9c7d7c0/1
    [25] https://advertise.tldr.tech/?utm_source=tldr&utm_medium=newsletter&utm_campaign=advertisecta
    [26] https://jobs.ashbyhq.com/tldr.tech
    [27] https://jobs.ashbyhq.com/tldr.tech/c227b917-a6a4-40ce-8950-d3e165357871 [28] https://www.linkedin.com/feed/update/urn:li:activity:7401699691039830016/ [29] https://twitter.com/tldrdan
    [30] https://twitter.com/SteveFlanders22
    [31] https://tldr.tech/tech/manage?email=tldrnewsletter%40synchro.net
    [32] https://a.tldrnewsletter.com/unsubscribe?ep=1&l=cfa2d55a-b7be-11e8-a3c9-06b79b628af2&lc=265280dc-d4e0-11f0-a661-1b9fa34a893d&p=0336dfc8-a6ac-11f1-87b3-2120317ce15f&pt=campaign&pv=4&spa=1788343235&t=1788345442&s=805bb951a1033093c133c109946a570a5f8c14e5f79fe842f6f2b9e7b68e7764

    ---
    þ Synchronet þ Vertrauen þ Home of Synchronet þ [vert/cvs/bbs].synchro.net
  • Who's Online

  • Recent Visitors

    • Len Bisop
      Wed Aug 12 02:34:52 2026
      from Toledo,oh via SSH
    • Guest
      Tue Aug 11 14:46:32 2026
      from Test via Telnet
    • Guest
      Sun Aug 9 23:07:37 2026
      from Butt, Bt via Telnet
    • Guest
      Sun Aug 9 16:35:02 2026
      from C via Telnet
  • System Info

    Sysop: RunStop
    Location: Huntsville, AL
    Users: 25
    Nodes: 4 (0 / 4)
    Uptime: 19:04:12
    Calls: 288
    Files: 8
    Messages: 32,988

© The C64 Pub, 2026