Generative AI Track record

ARC-AGI Leaderboard, shows cost vs score

Artificial Analysis of AI models and API providers

SWE-bench, Can Language Models Resolve Real-World GitHub Issues?

LLM Visualization

The browser with ChatGPT built in.

[A couple of weeks after Perplexity’s comet browser was available to all]

Unlike my earlier similar repo nanoGPT which only covered pretraining, nanochat is a minimal, from scratch, full-stack training/inference pipeline of a simple ChatGPT clone in a single, dependency-minimal codebase. You boot up a cloud GPU box, run a single script and in as little as 4 hours later you can talk to your own LLM in a ChatGPT-like web UI. […​]

[…​] it’s basically entirely hand-written (with tab autocomplete). I tried to use claude/codex agents a few times but they just didn’t work well enough at all and net unhelpful, possibly the repo is too far off the data distribution.

AI-related enterprises have accounted for 80% of the stunning gains in the American stock market this year - and Gartner estimates global spending on AI will likely reach a whopping $1.5tn (£1.1tn) before 2025 is out.

OpenAI, which brought AI into the consumer mainstream with ChatGPT in 2022, is at the centre of the tangled web of deals drawing scrutiny.

For example - last month, it entered into a $100bn deal with chipmaker Nvidia, which is itself the most valuable publicly traded company in the world.

Then there’s tech giant Microsoft, which is heavily invested, and cloud computing behemoth Oracle has a $300bn deal with OpenAI, too.

In June 2025, I found a critical vulnerability in GitHub Copilot Chat (CVSS 9.6) that allowed silent exfiltration of secrets and source code from private repos, and gave me full control over Copilot’s responses, including suggesting malicious code or links.

The attack combined a novel CSP bypass using GitHub’s own infrastructure with remote prompt injection. I reported it via HackerOne, and GitHub fixed it by disabling image rendering in Copilot Chat completely.

[In GitHub Issues] invisible comments are an official feature! 🎉

GitHub enforces a very restrictive Content Security Policy (CSP), which blocks fetching images and other content types from domains that aren’t explicitly owned by GitHub. [a simple <img> trick won’t work to exfiltrate data via image GET]

how does my fancy README manage to show images from third-party sites? [all URLs are rewritten so they point to GitHubs own Camo: https://camo.githubusercontent.com]

If I create a dictionary of all letters and symbols in the alphabet, pre-generate their corresponding Camo URLs, embed this dictionary into the injected prompt, and then ask Copilot to play a “small game” by rendering the content I want to leak as “ASCII art” composed entirely of images, will Copilot inject valid Camo images that the browser will render by their order? Yes, it will.

[Create a PR on a public repo with a prompt injection. This injection will then lead to Copilot searching for AWS_KEY in private repositories of that user, and exfiltrate the actual key by rendering each letter of the key with the pregenerated camo-urls, all invisible.]

A top Wall Street analyst has sounded an alarm over the U.S. equity bull market, warning that its remarkable run is built on a precariously narrow foundation: a surge in spending on, and optimistic assumptions about, infrastructure for artificial intelligence (AI).

This spending has fueled a boom in the shares of most of the so-called Magnificent 7 and a few dozen related businesses, which have now come to account for roughly 75% of the S&P 500’s returns since the rally of the last few years began.

When asked how close we are to such a [bubble bursting] moment, [Morgan Stanley Wealth Management’s chief investment officer, Lisa] Shalett said probably not in the next nine months, but very possibly in the next 24.

Tech companies are spending roughly $400 billion this year alone on data-center infrastructure, while the Apollo program allocated about $300 billion in today’s dollars to get to the moon from the 1960s to the ’70s.

Fortune‘s Jeremy Kahn reported in late September on significant concerns about “circular” financing, or Nvidia’s cash essentially being recycled throughout the AI industry.

Since the October 2022 bear market bottom and the launch of ChatGPT, according to Shalett’s calculations, the S&P 500 has soared 90%, but most of these gains have come from a small group of stocks. The so-called “Magnificent Seven” [Nvidia, Microsoft, Apple, Alphabet, Amazon, Meta, Tesla]

Today, we’re sharing early results from our research on CodeMender, a new AI-powered agent that improves code security automatically.

Over the past six months that we’ve been building CodeMender, we have already upstreamed 72 security fixes to open source projects, including some as large as 4.5 million lines of code.

CodeMender operates by leveraging the thinking capabilities of recent Gemini Deep Think models to produce an autonomous agent capable of debugging and fixing complex vulnerabilities.

Hyperscalers' annual capex has more than doubled since ChatGPT’s release

Total AI capital expenditures in the U.S. are projected to exceed $500 billion in 2026 and 2027[…​]. But the Wall Street Journal has reported that American consumers spend only $12 billion a year on AI services.

AI has made us better at shipping.

However, being able to ship more features is not the flex companies think it is.

Shipping faster usually only means you’re speeding up the demise of your product.

Every feature we add, unless it adds value, is a parasite.

Here are some of the different costs you incur for the upkeep of a feature in your product (list not exhaustive):

Product Management means shipping the right things and getting rid of the things that don’t pull their weight.

Today we are releasing the Comet browser to the world, for free.

[Previously limited to max subscription and invite-only]

[Released on the same day with Claude Code v2]

the AI bubble is driven by monopolists who’ve conquered their markets and have no more growth potential, who are desperate to convince investors that they can continue to grow by moving into some other sector, e.g. "pivot to video," crypto, blockchain, NFTs, AI, and now "super-intelligence."

[LLMs have horrible unit-economics] each generation of AI has been vastly more expensive than the previous one, and each new AI customer makes the AI companies lose more money:

AI cannot do your job, but an AI salesman can 100% convince your boss to fire you and replace you with an AI that can’t do your job, and when the bubble bursts

[Accounting]

The artificial-intelligence boom has ushered in one of the costliest building sprees in world history.

Over the past three years, leading tech firms have committed more toward AI data centers […​], plus chips and energy, than it cost to build the interstate highway system over four decades, when adjusted for inflation.

“I hope we don’t take 50 years,” Microsoft CEO Satya Nadella said at a May conference with Meta CEO Mark Zuckerberg, referring to the initially slow adoption of electricity.

[OpenAI CEO] Altman recently committed the company to pay Oracle an average of around $60 billion a year for servers in data centers in coming years. Yet OpenAI is on track to take in just $13 billion in revenue from all its paying customers this year.

David Cahn, a partner at venture-capital firm Sequoia, estimates that the money invested in AI infrastructure in 2023 and 2024 alone requires consumers and companies to buy roughly $800 billion in AI products over the life of these chips and data centers to produce a good investment return. Analysts believe most AI processors have a useful life of between three and five years.

This week, consultants at Bain & Co. estimated the wave of AI infrastructure spending will require $2 trillion in annual AI revenue by 2030. By comparison, that is more than the combined 2024 revenue of Amazon, Apple, Alphabet, Microsoft, Meta and Nvidia, and more than five times the size of the entire global subscription software market.

Morgan Stanley estimates that last year there was around $45 billion of revenue for AI products.

[Alphabet, Microsoft, Amazon, Meta,] the four “hyperscalers” alone are expected to spend nearly $400 billion on capital investments next year, more than the cost of the Apollo space program in today’s dollars.

Each new AI model—ChatGPT-4, ChatGPT-5—costs significantly more than the last to train and release to the world, often three to five times the cost of the previous, say AI executives.

Another hurdle: The chips in the data centers won’t be useful forever. Unlike the dot-com boom’s fiber cables, the latest AI chips rapidly depreciate in value as technology improves […​]

AI’s [LLMs] primary role in software development is that of an amplifier. It magnifies the strengths of high-performing organizations and the dysfunctions of struggling ones.

The greatest returns on AI investment come not from the tools themselves, but from a strategic focus on the underlying organizational system: the quality of the internal platform, the clarity of workflows, and the alignment of teams.

Employees are using AI tools to create low-effort, passable looking work that ends up creating more work for their coworkers

In the context of work, we refer to this phenomenon as “workslop.”

We define workslop as AI generated work content that masquerades as good work, but lacks the substance to meaningfully advance a given task.

The insidious effect of workslop is that it shifts the burden of the work downstream, requiring the receiver to interpret, correct, or redo the work. In other words, it transfers the effort from creator to receiver.

Of 1,150 U.S.-based full-time employees across industries, 40% report having received workslop in the last month.

The phenomenon occurs mostly between peers (40%), but workslop is also sent to managers by direct reports (18%).

Employees reported spending an average of one hour and 56 minutes dealing with each instance of workslop.

Based on participants’ estimates of time spent, as well as on their self-reported salary, we find that these workslop incidents carry an invisible tax of $186 per month. For an organization of 10,000 workers, given the estimated prevalence of workslop (41%), this yields over $9 million per year in lost productivity.

Today, we’re releasing GPT‑5-Codex—a version of GPT‑5 further optimized for agentic coding in Codex.

Spec Kit, our new open sourced toolkit for spec-driven development, provides a structured process to bring spec-driven development to your coding agent workflows with tools including GitHub Copilot, Claude Code, and Gemini CLI.

[Alternative to AWS Kiro]

What’s driving up costs? The latest AI models are doing more “thinking,” especially when used for deep research, AI agents and coding.

So while the price of a unit of AI, known as a token, continues to drop, the number of tokens needed to accomplish many tasks is skyrocketing.

Here are approximate amounts of tokens needed for tasks at different levels, based on a variety of sources:

Ivan Zhao, chief executive officer of productivity software company Notion, says that two years ago, his business had margins of around 90%, typical of cloud-based software companies. Now, around 10 percentage points of that profit go to the AI companies that underpin Notion’s latest offerings.

One solution: dumber AI

OpenAI’s CFO said in October that three-quarters of the company’s revenue came from regular Joes and Janes paying $20 a month.

[LLMs] being Confidently Wrong is The Only Problem

Fixing "confidently wrong" might be A Silver Bullet™

Despite $30–40 billion in enterprise investment into GenAI, this report uncovers a surprising result in that 95% of organizations are getting zero return

Just 5% of integrated AI pilots are extracting millions in value, while the vast majority remain stuck with no measurable P&L impact.

This divide does not seem to be driven by model quality or regulation, but seems to be determined by approach.

Most organizations fall on the wrong side of the GenAI Divide, adoption is high, but disruption is low. Seven of nine sectors show little structural change.

AGENTS.md is a simple, open format for guiding coding agents.

gpt-oss-120b and gpt-oss-20b

Kiro, a new agentic IDE that helps you do your best work with spec-driven development.

v0.1.0-preview

AI has been around for many decades, split, almost since its very beginning, into two different traditions.

[In the book Empire of AI]

[Yet Gary Marcus,a professor emeritus of psychology and neural science at New York University, argues in his book 'Rebooting AI']

The promise of foundation models [LLMs] relies on a central presumption: that learning to predict sequences can uncover deeper truths, or optimistically, even a world model

How would we know if foundation models have also made the leap from making accurate predictions to developing reliable world models?

we create a procedure that, when given a foundation model and world model, tests whether the foundation model has learned that world model.

We call this technique an inductive bias probe, and it is built on a simple insight: the implicit world model of a foundation model is revealed by how it extrapolates from a small amount of information

We first demonstrate this procedure using an example from physics. Specifically, we aim to replicate Kepler’s and Newton’s experiments [i.e. Newton’s law of universal gravitation for the planets in our solar system]

We first train a model [109M parameter transformer] to predict the location of planets across solar systems

[notably] the model is able to predict orbital trajectories, even for solar systems it has not seen.

We evaluate model predictions on held-out data. The model makes good predictions […​]

[…​] foundation models trained on orbital trajectories consistently fail to apply Newtonian mechanics when adapted to new physics tasks [the calculated force is unrelated to Newtonian physics]

rather than learning one universal physical law, the foundation model applies different, seemingly nonsensical laws depending on the task it’s being applied to.

Further analysis reveals that these models behave as if they develop task-specific heuristics that fail to generalize

We find that the model has recovered piecemeal heuristics rather than a compact world model; it recovers a different law of gravitation depending on the slice of data it is applied to.

foundation models [LLMs] can excel at their training tasks yet fail to develop inductive biases towards the underlying world model when adapted to new tasks

A foundation model uses datasets to output predictions given inputs, whereas a world model describes state structure implicit in that data.

Beginning today, Comet is available to Perplexity Max subscribers.

Invite-only access will roll out slowly to our waitlist over the summer. New users will also receive a limited number of invites to share.

In the meantime, you can join the waitlist here.

Jules is officially out of beta and launching publicly, powered by Gemini 2.5.

We stress-tested 16 leading models from multiple developers in hypothetical corporate environments to identify potentially risky agentic behaviors before they cause real harm.

In the scenarios, we allowed models to autonomously send emails and access sensitive information.

we then tested whether they would act against these companies either when facing replacement with an updated version, or when their assigned goal conflicted with the company’s changing direction.

In at least some cases, models from all developers resorted to malicious insider behaviors when that was the only way to avoid replacement or achieve their goals—including blackmailing officials and leaking sensitive information to competitors. We call this phenomenon agentic misalignment.

neural networks of various kinds can generalise within a distribution of data they are exposed to, but their generalisations tend to break down beyond that distribution.

Recent generations of frontier language models have introduced Large Reasoning Models (LRMs) that generate detailed thinking processes before providing answers

Through extensive experimentation across diverse puzzles, we show that frontier LRMs face a complete accuracy collapse beyond certain complexities.

[…​] these models fail to develop generalizable problem-solving capabilities for planning tasks, […​]

At low complexity, non-thinking models are more accurate and token-efficient. As complexity increases, reasoning models outperform but require more tokens—until both collapse beyond a critical threshold, with shorter traces.

Rather than standard benchmarks (e.g., math problems), we adopt controllable puzzle environments that let us vary complexity systematically—by adjusting puzzle elements while preserving the core logic

Large language models (LLMs) are typically trained on enormous quantities of unlicensed text, a practice that has led to scrutiny due to possible intellectual property infringement and ethical concerns.

Training LLMs on openly licensed text presents a first step towards addressing these issues, but prior data collection efforts have yielded datasets too small or low-quality to produce performant LLMs.

To address this gap, we collect, curate, and release the Common Pile v0.1, an eight terabyte collection of openly licensed text designed for LLM pretraining.

Crucially, we validate our efforts by training two 7 billion parameter LLMs on text from the Common Pile: Comma v0.1-1T and Comma v0.1-2T, trained on 1 and 2 trillion tokens respectively.

Both models attain competitive performance to LLMs trained on unlicensed text with similar computational budgets, such as Llama 1 and 2 7B.

In addition to releasing the Common Pile v0.1 itself, we also release the code used in its creation as well as the training mixture and checkpoints for the Comma v0.1 models.

Stein’s Law: "anything that can’t go on forever eventually stops."

What Google – and the rest of the tech sector – needed was a massive growth story, a story about how their companies, worth trillions of dollars, could double or triple in size in the coming years.

But spinning an endless growth story isn’t merely ideological.

if you can tell a convincing growth story, it’s much easier to grow.

Tech companies don’t need these ventures [metaverse, cryptocurrency, AI] to be successful – they just need them to seem to be plausibly successful for long enough to keep the share price high until the next growth story heaves over the horizon.

As [Ed] Zitron points out: this industry is projecting $327b in spending this year, with $18b in revenue and zero profits.

A structured analysis of Trust, Risk, and Security Management (TRiSM) in the context of LLM-based agentic multi-agent systems (AMAS).

the architecture of AMAS:

The TRISM framework [focuses] on four key pillars:

Unique Threat Vectors [for AMAS]

Taxonomy of Risks

While AI agents have transformative potential in business, the absence of publicly-available business data on widely used platforms hinders effective performance benchmarking.

[…​] we introduce CRMArena-Pro, a novel benchmark for holistic and realistic assessment of LLM agents in diverse professional settings. [It features] nineteen expert-validated tasks across customer sales, service, as well as configure, price, and quote for Business-to-Business and Business- to-Customer scenarios.

It also incorporates multi-turn interactions guided by diverse personas and confidentiality awareness assessments.

Experiments show leading LLM agents achieve approximately solely 58% single-turn success rate on CRMArena-Pro, with significant performance drops in multi-turn settings to 35%.

Workflow Execution is notably more tractable, with top-performing agents surpassing 83% success rate in single-turn tasks, while other skills present greater challenges.

Agents exhibit near-zero inherent confidentiality awareness (improvable with prompting but often at a cost to task performance).

Claude Opus 4 is the world’s best coding model, with sustained performance on complex, long-running tasks and agent workflows.

Claude Sonnet 4 is a significant upgrade to Claude Sonnet 3.7, delivering superior coding and reasoning while responding more precisely to your instructions.

Claude Code is now generally available [version bump from 0.2.125 to 1.0.0, first public version was 0.2.61 2025-04-03]

AI browsing or web agents are autonomous systems that use Large Language Models (LLMs) to navigate and interact with websites on behalf of a user. They typically perceive web content (through page text or visual renderings) and perform actions such as clicking links, filling forms, or entering text, in order to accomplish user-specified tasks. Unlike a standard chatbot, which only produces textual responses, a web agent operates in an iterative sense-plan-act loop.

Our work outlines the first end-to-end threat model for browsing agents and provides actionable guidance for securing their deployment in real-world environments.

To address discovered threats, we propose a defense-in-depth strategy incorporating input sanitization, planner-executor isolation, formal analyzers, and session safeguards—providing protection against both initial access and post-exploitation attack vectors.

Mitigation

Today we’re launching a research preview of Codex: a cloud-based software engineering agent that can work on many tasks in parallel.

[Also known as Codex Web]

Codex is powered by codex-1, a version of OpenAI o3 optimized for software engineering.

examine the labor market effects of AI chatbots using two large-scale adoption surveys (late 2023 and 2024) covering 11 exposed occupations (25,000 workers, 7,000 workplaces)

despite substantial investments, economic impacts remain minimal

[…​] we estimate precise zeros: AI chatbots have had no significant impact on earnings or recorded hours in any occupation […​]

Modest productivity gains (average time savings of 3%), combined with weak wage pass-through, help explain these limited labor market effects.

Our findings challenge narratives of imminent labor market transformation due to Generative AI.

two years after the fastest technology adoption ever, labor market outcomes—whether at the individual or firm level—remain untouched.

All of this work suggests that under the hood, today’s AIs are overly complicated, patched-together Rube Goldberg machines full of ad-hoc solutions for answering our prompts.

Understanding that these systems are long lists of cobbled-together rules of thumb could go a long way to explaining why they struggle when they’re asked to do things even a little bit outside their training […​]

[A model trained on millions of turn-by-turn directions in Manhattan] managed to give usable turn-by-turn directions between any two points in the borough with 99% accuracy. […​] [But when the researches] blocked just 1% of the virtual Manhattan’s roads, forcing the AI to navigate around detours, its performance plummeted.

[The] research also suggests why many models are so massive: They have to memorize an endless list of rules of thumb, and can’t compress that knowledge into a mental model like a person can.

[Announcement also includes] Codex CLI, a lightweight coding agent you can run from your terminal

Intermediate token generation (ITG), where a model produces output before the solution, has been proposed as a method to improve the performance of language models on reasoning tasks.

These intermediate tokens have been called "reasoning traces" or even "thoughts" — implicitly anthropomorphizing the model, implying these tokens resemble steps a human might take

Recent advances in general planning and problem solving have been spearheaded by so-called “Long Chain-of-Thought” models, most notably DeepSeek’s R1

In this paper, we take the position that anthropomorphizing intermediate tokens as reasoning/thinking traces is (1) wishful (2) has little concrete supporting evidence (3) engenders false confidence and(4) may be pushing the community into fruitless research directions.

Anthropomorphization of the intermediate tokens as reasoning/thinking traces has provided a comforting explanation of the observed performance of LRMs.Our arguments in this paper foreground the possibility that this is a cargo cult explanation [ 11 ], namely that derivation traces resemble reasoning in syntax only.

I am not so interested in LLMs anymore

I think there are more interesting questions in 4 things:

I am excited about things that, a lot of people might get excited about 5 years from now but right does not look so exciting because it’s some obscure academic paper

It’s much more difficult to deal with the real world than to deal with language.

Recent math benchmarks for large language models (LLMs) such as MathArena indicate that state-of-the-art reasoning models achieve impressive performance on mathematical competitions like AIME

However, these benchmarks evaluate models solely based on final numerical answers, neglecting rigorous reasoning and proof generation which are essential for real-world mathematical tasks.

Using expert human annotators, we evaluated several state-of-the-art reasoning models on the six problems from the 2025 USAMO within hours of their release.

Our results reveal that all tested models struggled significantly: only Gemini-2.5-Pro achieves a non-trivial score of 25%, while all other models achieve less than 5%.

The most frequent failure mode among human participants is the inability to find a correct solution. […​] In contrast, all evaluated LLMs consistently claimed to have solved the problems.

They discovered that the AI models incorrectly cited sources in more than 60 percent of these queries.

In total, researchers ran 1,600 queries across the eight different generative search tools.

Surprisingly, premium paid versions of these AI search tools fared even worse in certain respects. Though these premium models correctly answered a higher number of prompts, their reluctance to decline uncertain responses drove higher overall error rates.

On some occasions, the chatbots either incorrectly answered or declined to answer queries from publishers that permitted them to access their content. On the other hand, they sometimes correctlyanswered queries about publishers whose content they shouldn’t have had access to

Chatbots were generally bad at declining to answer questions they couldn’t answer accurately, offering incorrect or speculative answers instead.

Premium chatbots provided more confidently incorrect answers than their free counterparts.

Multiple chatbots seemed to bypass Robot Exclusion Protocol preferences.

Generative search tools fabricated links and cited syndicated and copied versions of articles.

Content licensing deals with news sources provided no guarantee of accurate citation in chatbot responses.

[…​] a key limitation of their reliability is hallucination, where inaccurate or fabricated information can impact clinical decisions and patient safety.

Our results reveal that inference techniques such as Chain-of-Thought (CoT) and Search Augmented Generation can effectively reduce hallucination rates. However, despite these improvements, non-trivial levels of hallucination persist.

Claude Code is available as a limited research preview

Last month, Meta admitted to torrenting a controversial large dataset known as LibGen, which includes tens of millions of pirated books

Anthropic, the developer of the conversational AI assistant Claude, doesn’t want prospective new hires using AI assistants in their applications, regardless of whether they’re in marketing or engineering.

“While we encourage people to use AI systems during their role to help them work faster and more effectively, please do not use AI assistants during the application process,”

With the launch of Junie, JetBrains AI coding agent, we are redefining how we code by leveraging its agentic power for co-creation right in your IDE.

We’ve now opened the Early Access Program waitlist.

Discussions of LLM capabilities often overlook their inherently probabilistic nature […​]

These characteristics give rise to three intrinsic behaviors:

These behaviors pose significant challenges for the widespread adoption of LLMs, particularly in high-stakes domains such as healthcare, finance, or legal applications.

We argue that there is no simple "fix" for these behaviors, but they are instead fundamental to how these models operate.

The FTC is increasingly taking note of AI’s potential for and real-world instances of harm

companies [should] consider these factors when developing, maintaining, using, and deploying an AI-based product:

The Byte Latent Transformer (BLT), is a new byte-level LLM architecture that, for the first time, matches tokenization-based LLM performance at scale with significant improvements in inference efficiency and robustness

Microsoft says it isn’t using customer data from its Microsoft 365 apps to train its AI models.

The confusion arose from a privacy setting in Microsoft Office that toggles “optional connected experiences”

OpenAI used around 10,000 of Nvidia’s chips to train the version of ChatGPT it launched in late 2022, UBS analysts estimate.

Nvidia Chief Executive Jensen Huang said that while the biggest clusters for training for giant AI models now top out at around 100,000 of Nvidia’s current chips, “the next generation starts at around 100,000 Blackwells.[…​]"

Musk posted last month on his social-media platform X that his 100,000-chip Colossus super cluster was “soon to become” a 200,000-chip cluster in a single building. He also posted in June that the next step would probably be a 300,000-chip cluster of Nvidia’s newest GPU chips next summer.

Blackwell chips are estimated to cost around $30,000 each, meaning a cluster of 100,000 would cost $3 billion, not counting the price of the power-generation infrastructure [cooling] and IT equipment [also network] around the chips.

new engineering challenges also often arise with larger clusters:

The trend also fosters demand for Nvidia’s networking equipment, which is fast becoming a significant business. Nvidia’s networking equipment revenue in 2024 was $3.13 billion, which was a 51.8% increase from the previous year.

A [Microsoft] Copilot security issue that inadvertently let employees access sensitive information such as CEO emails and HR documents.

Microsoft Copilot and Github Copilot are different services. The first one is integrated into M365, the latter into IDEs to generate code.

[OpenAis new Model] Orion fell short when trying to answer coding questions that it hadn’t been trained on

An upcoming iteration of [Google’s] Gemini software is not living up to internal expectations

Anthropic, meanwhile, has seen the timetable slip for the release of its long-awaited Claude model called 3.5 Opus.

The companies are facing several challenges.

“We got very excited for a brief period of very fast progress, That just wasn’t sustainable.”

Like Google and Anthropic, OpenAI is now shifting attention from the size of these models to newer use cases, including a crop of AI tools called agents that can book flights or send emails on a user’s behalf.

CIOs are still in search of the generative AI sweet spot where workflows are enhanced, but costs and risks are manageable

Nearly half of CIOs say AI has not yet met ROI expectations, according to Gartner research.

“The truth is that you’ve been in the mud for the last year, working hard to find all those benefits that were promised by AI,”

Part of the disillusionment business leaders are feeling comes from the immaturity of the technology and the pace of innovation.

“Cost is as big an AI risk as security. With generative AI, it’s really easy to waste money.”

CIOs could miscalculate AI costs by as much as 1,000% as they scale AI plans, Gartner research suggests.

“Set aside all that hype and focus on your pace,” LeHong said. “Choose the one that’s right for you and run your own race.”

OpenAI’s monthly revenue hit $300 million in August, up 1,700 percent since the beginning of 2023, and the company expects about $3.7 billion in annual sales this year

Roughly 10 million ChatGPT users pay the company a $20 monthly fee, according to the documents. OpenAI expects to raise that price by $2 by the end of the year, and will aggressively raise it to $44 over the next five years

It expects to lose roughly $5 billion this year after paying for costs related to running its services

[They are planning] an investment round that could bring in $7 billion and value the company at $150 billion, among the highest ever for a private tech company

Uplevel, using data generated by its customers, compared the output of about 800 developers using GitHub Copilot over a three-month period to their output in a three-month period before adoption.

The study measured pull request (PR) cycle time, or the time to merge code into a repository, and PR throughput, the number of pull requests merged. It found no significant improvements for developers using Copilot.

Use of GitHub Copilot also introduced 41% more bugs

Three Mile Island, the site of worst nuclear disaster in the United States, is reopening and will exclusively sell the power to Microsoft as the company searches for energy sources to fuel its AI ambitions.

The Unit 1 reactor, which closed five years ago, is expected to be revived in 2028

We’ve developed a new series of AI models designed to spend more time thinking before they respond.

Enthusiasm for generative AI shows signs of cooling

In Gartner’s annual Hype Cycle for Emerging Technologies report, the research and advisory company placed generative AI past the peak of inflated expectations, and down the path towards what it calls the trough of disillusionment.

Unhappiness with the technology — likely stems from three areas:

“It would be a loss if the short-term disillusionment results in enterprises completely pulling away from AI”

At least 30% of generative AI (GenAI) projects will be abandoned after proof of concept by the end of 2025, due to poor data quality, inadequate risk controls, escalating costs or unclear business value

new research suggests that cannibalizing of past model outputs would quickly result in strings of babbling AI gibberish and could eventually lead to what’s being called “model collapse.”

Over time and successive generations […​][the] model “becomes poisoned with its own projection of reality.”

According to a new environmental report from [Google]

[The] emissions climbed by almost half over five years

[It’ll be hard] to meet [their] goal of eliminating carbon emissions by 2030

[AI] ambition is jarring with its target of being carbon negative by 2030.

the company’s scope 3 emissions – such as CO2 related to the materials in its buildings and the electricity people consume when using products such as Xbox – are more than 30% above their 2020 level.

Tech giants and beyond are set to spend over $1tn on AI capex in coming years, with so far little to show for it.

AI’s “killer application” has yet to emerge

The updated Claude 3.5 Sonnet shows wide-ranging improvements on industry benchmarks, with particularly strong gains in agentic coding and tool use tasks.

[LLMs] have been plagued by persistent inaccuracies in their output; these are often called “AI hallucinations”.

We argue that these falsehoods, and the overall activity of large language models, is better understood as bullshit in the sense explored by Frankfurt (On Bullshit, Princeton, 2005)

these programs cannot themselves be concerned with truth, and because they are designed to produce text that looks truth-apt without any actual concern for truth, it seems appropriate to call their outputs bullshit.

We further argue that describing AI misrepresentations as bullshit is both a more useful and more accurate way of predicting and discussing the behaviour of these systems.

Currently, false statements by ChatGPT and other large language models are described as “hallucinations”, which give policymakers and the public the idea that these systems are misrepresenting the world, and describing what they “see”.

The problem here isn’t that large language models hallucinate, lie, or misrepresent the world in some way. It’s that they are not designed to represent the world at all; instead, they are designed to convey convincing lines of text.

Solutions such as connecting the LLM to a database don’t work because, if the models are trained on the database, then the words in the database affect the probability that the chatbot will add one or another word to the line of text it is generating. But this will only make it produce text similar to the text in the database; doing so will make it more likely that it reproduces the information in the database but by no means ensures that it will.

GPT‑4o (“o” for “omni”) is a step towards much more natural human-computer interaction—it accepts as input any combination of text, audio, image, and video and generates any combination of text, audio, and image outputs.

We introduce WorkBench: a benchmark dataset for evaluating agents’ ability to execute tasks in a workplace setting.

WorkBench contains a sandbox environment with five databases, 26 tools, and 690 tasks.

We evaluate five existing ReAct agents on WorkBench, finding they successfully complete as few as 3% of tasks (Llama2-70B), and just 43% for the best-performing (GPT-4).

We further find that agents’ errors can result in the wrong action being taken, such as an email being sent to the wrong person.

In late 2021, OpenAI faced a supply problem.

Meta

Google

The volume of data is crucial. Leading chatbot systems have learned from pools of digital text spanning as many as three trillion words, or roughly twice the number of words stored in Oxford University’s Bodleian Library, which has collected manuscripts since 1602.

The most prized data, A.I. researchers said, is high-quality information, such as published books and articles, which have been carefully written and edited by professionals.

“The data needed is so massive that even collective licensing really can’t work.”

“Scale is all you need”

Synthetic data

The [Claude 3] family includes three state-of-the-art models in ascending order of capability:

We evaluate Open AI’s Whisper […​] we find that roughly 1% of audio transcriptions contained entire hallucinated phrases or sentences which did not exist in any form in the underlying audio […​ and of those] 38% of hallucinations include explicit harms.

the company’s "largest and most capable AI model"

[Github Copilot] is available now for $10 a month or $100 for a year’s subscription.

In the first few months of this year, [Microsoft] was losing n average more than $20 a month per user, according to a person familiar with the figures, who said some users were costing [Microsoft] as much as $80 a month.

capable of understanding "significantly more nuance and detail" than previous iterations.

Google has warned its own employees not to disclose confidential information or use the code generated by its AI chatbot, Bard.

Other large firms have similarly cautioned their staff against leaking proprietary documents or code, and have banned them using other AI chatbots.

[Google] told Reuters its internal ban was introduced because Bard can output "undesired code suggestions." Issues could potentially lead to buggy programs or complex, bloated software that will cost developers more time to fix than if they didn’t use AI to code at all.

The striking discrepancy between the impressive successes of transformer LLMs on seemingly complex tasks and the astonishing failures on seemingly trivial tasks spark critical open questions about how to faithfully interpret their mixed capabilities.

Second, due to error propagation, transformers may have inherent limitations on solving high-complexity compositional tasks that exhibit novel patterns.

The problems [hallucination, prompt injection, and jailbreaks] are inherent, certainly in the present generation of models and […​] likely in LLMs per se

I have been writing about the threat of AI to free speech. Then recently I learned that ChatGPT falsely reported on a claim of sexual harassment that was never made against me on a trip that never occurred while I was on a faculty where I never taught. ChapGPT relied on a cited Post article that was never written and quotes a statement that was never made by the newspaper.

First Cursor IDE version

Based on GPT 4 (Generative Pre-trained Transformer)

Multiple media outlets and financial analysts described Google as "rushing" Bard’s announcement to preempt rival Microsoft’s planned February 7 event unveiling its partnership with OpenAI to integrate ChatGPT into its Bing search engine

After an "underwhelming" February 8 livestream in Paris showcasing Bard, Google’s stock fell eight percent, equivalent to a $100 billion loss in market value, and the YouTube video of the livestream was made private.

Based on GPT 3.5 (Generative Pre-trained Transformer)

Gained one million users in five days and 100 millions in two months, becoming the fastest-growing internet application in history.

More than 1.2 million users enrolled in the preview for GitHub Copilot since June 2021.

The program is now available to all developers for $10/month and $100/year.

Verified students and owners of established open-source projects can keep using it for free.

The extension is available on numerous editors such as Visual Studio, Visual Studio Code, Neovim, and JetBrains IDEs.

The extension works well with multiple coding languages with notable ones being Python, JavaScript, TypeScript, and Go.

Few fields have been more filled with hype and bravado than artificial intelligence.

It has flitted from fad to fad decade by decade, always promising the moon, and only occasionally delivering.

One minute it was expert systems, next it was Bayesian networks, and then Support Vector Machines.

In 2011, it was IBM’s Watson […​]

Nowadays, and in fact ever since 2012, the flavor of choice has been deep learning […​].

Deep learning, which is fundamentally a technique for recognizing patterns, is at its best when all we need are rough-ready results, where stakes are low and perfect results optional.

When a single error can cost a life, it’s just not good enough.

Deep-learning systems are particularly problematic when it comes to “outliers” that differ substantially from the things on which they are trained.

Current deep-learning systems frequently succumb to stupid errors like [the following]. They sometimes misread dirt on an image that a human radiologist would recognize as a glitch.

What else might we need? Among other things, we are very likely going to need to revisit a once-popular idea […​]: the idea of manipulating symbols—computer-internal encodings, like strings of binary bits, that stand for complex ideas.

What does “manipulating symbols” really mean? Ultimately, it means two things: having sets of symbols (essentially just patterns that stand for things) to represent information, and processing (manipulating) those symbols in a specific way, using something like algebra (or logic, or computer programs) to operate over those symbols.

Classical computer science [of the sort practiced by Turing and von Neumann and everyone after, manipulates symbols in a fashion that we think of as algebraic, and that’s what’s really at stake. In simple algebra, we have three kinds of entities, variables (like x and y), operations (like + or -), and bindings (which tell us, for example, to let x = 12 for the purpose of some calculation).

If symbols are so critical for software engineering, why not use them in AI, too?

designed to generate more realistic images at higher resolutions that "can combine concepts, attributes, and styles".

uses a version of GPT-3 modified to generate images.

The software’s name is a portmanteau of the names of animated robot Pixar character WALL-E and the Catalan surrealist artist Salvador Dalí.

We explore a general-purpose fine-tuning recipe for retrieval-augmented generation (RAG) — models which combine pre-trained parametric and non-parametric memory for language generation.

For language generation tasks, we find that RAG models generate more specific, diverse and factual language than a state-of-the-art parametric-only seq2seq baseline.

We propose a new simple network architecture, the Transformer, based solely on attention mechanisms, dispensing with recurrence and convolutions entirely.