Recommended Free Tools
Some links on this page are affiliate links: if you buy through them we may earn a commission, at no extra cost to you.
Gary Marcus argues that fluent large language models are not, by themselves, a reliable route to human-level artificial intelligence. He says current systems remain weak at common sense, causal understanding, grounding, and dependable reasoning, and favors combining machine learning with structured representations and verification. Marcus is a cognitive scientist, author, entrepreneur, and AI commentator; to understand both his technical critique and his policy proposals, start with Rebooting AI and Taming Silicon Valley, then follow his current essays at Marcus on AI.
Who is Gary Marcus?
Gary Marcus is a cognitive scientist, best-selling author, entrepreneur, and commentator on artificial intelligence. His work crosses psychology, neuroscience, genetics, linguistics, and AI. That range matters to his public arguments: he often approaches machine intelligence by asking what human cognition can tell us about the kinds of representations and capabilities an advanced system would need.
As an Amazon Associate I earn from qualifying purchases.
Outdated Drivers Are Slowing You Down
One free scan finds every outdated or missing driver and matches the right update for your exact hardware.Free scan · exact hardware matchPC Slower Than It Used to Be?
A free scan shows the junk files, broken settings and background clutter dragging Windows down - then fixes them in one click.Free scan · Windows 10 & 11Carnegie Council describes Marcus as founder and CEO of Robust.AI and as the former founder and CEO of Geometric Intelligence, which Uber acquired in 2016. Institutional biographies also describe him as an NYU professor emeritus. His career includes both academic work on the mind and entrepreneurship in AI, rather than only commentary on new chatbot releases.
Marcus began programming as a child, studied cognitive science at Hampshire College and MIT, and taught psychology and neural science at New York University. His institutional profiles present a through-line from studying human cognition to asking how machine intelligence might be built and evaluated. They also identify writing for general audiences, including contributions to The New Yorker and The New York Times.
#1 Best Overall
- Read Before You Buy — No Video Output: These adapters support charging and USB 2.0 data transfer, but cannot transmit video signals. Except for standard USB webcams (which use USB data only), they are not compatible with HDMI/DisplayPort cables, video-capable USB-C hubs, or docking stations with video output.
- Convert USB-A Ports to USB-C: Designed to connect USB-C earphones, cables, flash drives, card readers, and other USB-C accessories to standard USB-A ports. Plug-and-play with no drivers or software required.
- Aluminum Alloy Housing: Built with a sturdy aluminum alloy shell that aids in heat dissipation and protects against daily wear and scratches. Designed to maintain a stable and secure connection.
- Compact & Travel-Friendly: The ultra-compact design allows the adapter to stay plugged into your device without blocking adjacent ports or adding bulk, reducing wear and tear on your original USB ports.
- 12-Month Warranty: Backed by a 12-month manufacturer warranty for peace of mind. Designed to meet strict quality control standards for reliable everyday performance.
What does Gary Marcus argue about ChatGPT and large language models?
Marcus’s central criticism is that statistical systems can produce convincing language without reliably understanding or representing the world they describe. In his framework, fluency is not proof of common sense, causal knowledge, or robust reasoning. This distinction is important when evaluating a system: a plausible answer may still be ungrounded or wrong.
He points to recurring limitations in current deep-learning systems: weak common sense, limited causal understanding, insufficient explicit knowledge, fragile models of time and space, and unreliable reasoning. These are related but distinct concerns. A model might generate a coherent description of an event yet fail to track how the event unfolded, what caused it, or what would change if one condition were different.
That critique does not mean Marcus regards generative AI as useless. An IEEE Spectrum interview records his view that it can be useful for bounded purposes such as autocomplete, while arguing that its capabilities are overvalued when treated as a path to artificial general intelligence (AGI). The practical distinction is between using a system for a limited task where its output can be checked and treating fluent output as evidence of broad, dependable competence.
Marcus has put his distance from near-term AGI claims plainly: “I don’t think we’re anywhere near AGI.” In the same IEEE Spectrum interview, he said, “I suspect we need to take a broad view of how the human mind works if we’re to build really advanced AI.” Those statements capture two parts of his position: skepticism about claims that current generative systems are close to general intelligence, and an interest in drawing on a wider understanding of cognition to build more capable systems.
What does he think today’s AI systems are missing?
Marcus’s concerns are easier to assess when separated into the capabilities he says are weak and the consequences of those weaknesses. He is not simply objecting to a particular chatbot’s style or to the use of neural networks. His broader question is whether systems built primarily by learning statistical patterns can support the structured, dependable understanding that people expect from advanced intelligence.
Rank #2
- 5-in-1 USB-C Hub: Experience comprehensive connectivity featuring a Power Delivery input, two USB-A 2.0 ports, a USB-A 3.0 port, and an HDMI port. (Note: The USB-C power delivery input port is only for connecting an external wall charger to power your laptop and cannot power peripheral devices.)
- 90W Pass-Through Charging: Achieve optimal charging with 90W pass-through power to your laptop, supported by a total input of 100W, with the hub reserving 10W for operational efficiency. (Note: Wall charger not included.)
- Quick Data Transfers: Accelerate your productivity with rapid data transfers using a high-speed 5Gbps USB 3.0 port and two 480Mbps USB 2.0 ports.
- 4K HDMI Display: Enhance your visual experience with a hub capable of delivering 4K resolution at 30Hz in both mirror and extend modes. Please note that this hub is compatible with MacBook (macOS 12 and newer), Windows 10 and 11, ChromeOS, and laptops equipped with DP Alt Mode and Power Delivery. Note: This device is not compatible with Linux.
- What You Get: Anker USB-C Hub (5-in-1, 4K HDMI), welcome guide, 18-month warranty, and our friendly customer service.
Common sense and grounding
Common sense lets people interpret ordinary situations against background knowledge about objects, agents, and events. Grounding is the connection between a representation and the world it is meant to describe. Marcus argues that a system can imitate the form of a sensible answer without possessing a sufficiently reliable basis for it. For users, this is a reason to distinguish an answer that sounds natural from one that has been checked against dependable evidence.
The Tool Desk
Outbyte Driver Updater FREEScan for outdated or missing drivers - takes under a minuteDriver Scan →Outbyte PC Repair FREEClear out junk files and repair common Windows errorsFree Scan →Causality, time, and space
Causal understanding involves more than noticing that two things often appear together: it concerns how one event produces or changes another. Marcus also identifies robust representations of time and space as shortcomings. These matter when a task depends on tracking what happened first, where something is, or how a change in circumstances should affect a conclusion. A model may handle familiar language patterns while still failing to reason consistently through such relationships.
Explicit knowledge and reliable reasoning
Marcus’s preferred alternative gives explicit knowledge a role alongside learning. In his account, systems need more than the ability to generate likely continuations; they need ways to represent relevant facts and support dependable inferences. The emphasis on reliability also shifts attention from impressive demonstrations to whether a system remains sound across the cases in which people intend to use it.
What does Marcus propose instead of scaling alone?
Marcus favors hybrid, or neurosymbolic, systems that combine learning with structured representations and verification. The idea is not to discard learning, but to pair its ability to find patterns with explicit structures that can represent knowledge and help check conclusions. This addresses his concern that statistical fluency alone does not guarantee grounded or dependable behavior.
His approach also points toward a different way to judge progress. If the goal is trustworthy intelligence, evaluations should examine whether a system can handle common sense, causal relationships, time, space, and reasoning reliably—not only whether it can produce polished language. The available accounts establish those priorities but do not provide one definitive test or a settled engineering recipe. It is more accurate to describe neurosymbolic AI as the direction he advocates than as a finished solution that has already resolved the limitations he identifies.
Rank #3
- Sleek 7-in-1 USB-C Hub: Features an HDMI port, two USB-A 3.0 ports, and a USB-C data port, each providing 5Gbps transfer speeds. It also includes a USB-C PD input port for charging up to 100W and dual SD and TF card slots, all in a compact design.
- Flawless 4K@60Hz Video with HDMI: Delivers exceptional clarity and smoothness with its 4K@60Hz HDMI port, making it ideal for high-definition presentations and entertainment. (Note: Only the HDMI port supports video projection; the USB-C port is for data transfer only.)
- Double Up on Efficiency: The two USB-A 3.0 ports and a USB-C port support a fast 5Gbps data rate, significantly boosting your transfer speeds and improving productivity.
- Fast and Reliable 85W Charging: Offers high-capacity, speedy charging for laptops up to 85W, so you spend less time tethered to an outlet and more time being productive.
- What You Get: Anker USB-C Hub (7-in-1), welcome guide, 18-month warranty, and our friendly customer service.
For a reader deciding what his argument means in practice, the key question is not whether an AI system is neural or symbolic in name. It is whether the design and evaluation provide reliable support for the task at hand. Marcus’s critique is aimed at treating scale or fluent output as sufficient evidence that a system can understand, reason, or act safely across broad contexts.
Is Gary Marcus an AI skeptic or a doomer?
“AI skeptic” is a more useful shorthand for Marcus’s position than “doomer,” provided it is not taken to mean that he rejects every AI application. The accounts of his work describe both interest in beneficial uses and concern that generative systems are being overvalued, deployed without sufficient accountability, or treated as closer to general intelligence than he believes they are.
His skepticism focuses on capability claims and governance. He questions whether current systems have the grounding and reasoning needed for broad reliability, while raising risks such as misinformation, deepfakes, copyright disputes, opaque behavior, concentrated power, and weak accountability. Those are concerns about how systems are designed, represented, and governed, not a claim that every AI use is inherently harmful.
That distinction helps readers interpret his commentary. A criticism of an AGI forecast is not the same as a forecast that AI will inevitably cause catastrophe. And calling for stronger oversight does not imply opposition to useful applications. Marcus argues for democratic oversight and accountability while retaining space for beneficial AI.
What are his policy concerns?
Marcus’s policy argument is that technical capability alone does not decide whether AI serves the public. The institutions that control data, systems, and deployment also shape who benefits, who bears risks, and who can challenge harmful outcomes. His concerns include mass-produced misinformation, deepfake pornography, disputes over copyrighted training material, black-box behavior, concentrated corporate power, and technology companies’ ability to shape public policy.
Rank #4
- Dual Converters, Infinite Potential：Includes 2× USB C male to USB A female adapters and 2× USB A male to USB C female adapters. Perfect for a wide range of uses—tablets with Bluetooth keyboards, expand USB ports on macbook, and more. Two different converters for all your daily needs
- Next-Level 10Gbps & 3A Charging: No more slow 480Mbps, this usb to usb c adapter has a transfer speed of up to 10Gbps, allowing you to do more transferring in less time. This usb adapter fits both USB A and USB C charger, supporting up to 3A fast charging
- Upgraded Exquisite Craftsmanship: With an aluminum alloy housing and metal connector, the usbc to usb adapter is extremely durable and sturdy. Rigorously tested to withstand more than 10,000 times of plugging and unplugging, ensuring long-lasting performance
- Broad Compatible: The usb c to usb adapter widely supports all USB C/ USB A devices like laptops, tablets, cellphones, car chargers, and phone chargers. Such as compatible with MacBook Pro/Air 2023/2022, Thunderbolt 4/3 Devices,Apple MagSafe Watch 9/8/7/SE/Ultra, iPad Pro 2022/2021, Samsung Galaxy S23/S20/S10, and iPhone 17/16/15 Pro. Plug and play
- Please Note: To reach 10Gbps speed, keep the cable under 3.3 ft. For USB A Male to USB C adapters, try flipping the USB C connector. USB C Male to USB A adapters support bidirectional 10Gbps transfer within 3.3 ft
In Taming Silicon Valley: How We Can Ensure That AI Works for Us, Marcus makes the policy case most directly. MIT Press describes the book as addressing Big Tech’s influence and AI’s potential harms, and identifies a set of proposed tools: data rights, layered AI oversight, transparency, liability, and tax reform. The policy focus complements his technical critique. If systems are difficult to inspect or hold responsible, reliable performance is not only a matter of model design; it is also a question of oversight and accountability.
These proposals are presented as ways to retain beneficial uses while responding to the power and risks associated with AI. They should not be read as a promise that one policy measure can solve the technical problems Marcus identifies. Rather, the book connects governance questions—who sets rules, what information people control, and who is liable—to the broader aim of ensuring that AI works for people.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.Which Gary Marcus book or essay should you read first?
Choose a starting point based on whether you want the technical diagnosis, the governance argument, a shorter critique of deep learning, or current commentary. The works address related questions but differ in audience and purpose.
Free tools Windows power users keep installed
One-click scans. No signup required.
| Start here | Best for | What it covers |
|---|---|---|
| Rebooting AI: Building Artificial Intelligence We Can Trust, with Ernest Davis | Readers seeking an accessible introduction to Marcus’s technical critique | Missing common sense, causal models, and the design of more trustworthy systems. |
| Taming Silicon Valley: How We Can Ensure That AI Works for Us | Readers interested in governance and public policy | Data rights, oversight, transparency, liability, tax reform, and democratic accountability. |
| Deep Learning: A Critical Appraisal | Readers looking for a technical treatment | A critical examination of deep learning and the limitations Marcus believes it faces. |
| “Deep Learning Is Hitting a Wall” | Readers who want a general-audience version of the scaling critique | A broader account of his argument that deep learning alone may not deliver the capabilities people expect. |
| Marcus on AI | Readers who want ongoing commentary | Posts on LLM reliability, AGI claims, AI economics, safety, and policy. |
A practical reading sequence is to begin with Rebooting AI for the technical diagnosis, then move to Taming Silicon Valley for the policy response. Readers who want to examine the deep-learning critique in shorter or more technical form can add the essay or appraisal; those following current arguments can use the newsletter archive for dated posts.
What is Taming Silicon Valley?
Published by MIT Press on September 17, 2024, Taming Silicon Valley is Marcus’s clearest policy-focused book. MIT Press lists it at 248 pages and gives paperback ISBN 9780262551069 and eBook ISBN 9780262381567. The publisher’s listed paperback price is $18.95; price and availability can vary by seller and location, so check the publisher for current details.
Best Value
- 5-in-1 Connectivity: Equipped with a 4K HDMI port, a 5 Gbps USB-C data port, two 5 Gbps USB-A ports, and a USB C 100W PD-IN port. Note: The USB C 100W PD-IN port supports only charging and does not support data transfer devices such as headphones or speakers.
- Powerful Pass-Through Charging: Supports up to 85W pass-through charging so you can power up your laptop while you use the hub. Note: Pass-through charging requires a charger (not included). Note: To achieve full power for iPad, we recommend using a 45W wall charger.
- Transfer Files in Seconds: Move files to and from your laptop at speeds of up to 5 Gbps via the USB-C and USB-A data ports. Note: The USB C 5Gbps Data port does not support video output.
- HD Display: Connect to the HDMI port to stream or mirror content to an external monitor in resolutions of up to 4K@30Hz. Note: The USB-C ports do not support video output.
- What You Get: Anker 332 USB-C Hub (5-in-1), welcome guide, our worry-free 18-month warranty, and friendly customer service.
Where does Gary Marcus publish now?
Marcus on AI is his active publication site and archive. Its posts cover LLM reliability, claims about AGI, AI economics, safety, and policy; the homepage is the place to find recent writing, while the archive lets readers locate essays by date.
For stable reference material, MIT Press is the authoritative publisher page for Taming Silicon Valley, including its formats, publication date, ISBNs, and retailer information. IEEE Spectrum’s author page and interview offer accessible biographical background and a concise entry into Marcus’s views on generative AI. Readers looking for earlier writing can use his institutional biographies, which identify work for The New Yorker and The New York Times, and the Frontier Minds reading list, which points to his essays, his 2023 Senate testimony, and Rebooting AI.
Do these 3 things before closing this tab:
1Fix the driver behind crashes, sound loss and screen glitches2Repair Windows errors before they cause bigger problems3Scan for outdated or missing drivers - takes under a minuteThese sources serve different purposes: the newsletter follows his current arguments, the publisher describes the book, and the profile and interview provide context. When reading his changing views on new systems or claims, dated essays are useful because they show what he was responding to at the time.
Frequently Asked Questions
What are the ISBNs for Gary Marcus’s Taming Silicon Valley?
MIT Press lists paperback ISBN 9780262551069 and eBook ISBN 9780262381567. Those identifiers distinguish the two formats when searching a library or bookseller catalogue.
Where has Gary Marcus published general-audience writing?
Institutional biographies identify contributions to The New Yorker and The New York Times. Frontier Minds also points readers to his essay “Deep Learning Is Hitting a Wall” and his 2023 Senate testimony.
Did Gary Marcus found an AI company?
Yes. Carnegie Council describes him as founder and CEO of Robust.AI and as the former founder and CEO of Geometric Intelligence. Uber acquired Geometric Intelligence in 2016.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
