{"id":1940,"date":"2026-07-01T05:01:00","date_gmt":"2026-07-01T05:01:00","guid":{"rendered":"https:\/\/futurenews24.com\/index.php\/2026\/07\/01\/july-2026-newsletter\/"},"modified":"2026-07-06T06:59:09","modified_gmt":"2026-07-06T06:59:09","slug":"july-2026-newsletter","status":"publish","type":"post","link":"https:\/\/futurenews24.com\/index.php\/2026\/07\/01\/july-2026-newsletter\/","title":{"rendered":"July 2026 Publication – by Piers Stobbs"},"content":{"rendered":"


\n<\/p>\n

\n

Hello everyone-<\/p>\n

Fairly a month! World Cup kicking off, offers (or no offers) within the Straits of Hormuz, a possible new PM, and the most well liked days on file… positively time for a quiet learn within the shade with all of the goings on within the vast world of AI and Knowledge Science. A number of nice content material beneath and I actually encourage you to learn on, however listed here are the edited highlights in case you are quick for time!<\/p>\n

Following is the most recent version of our Knowledge Science and AI publication. Hopefully some fascinating subjects and titbits to feed your information science curiosity. NOTE: If the e-mail doesn\u2019t show correctly (it’s fairly lengthy\u2026) click on on the \u201cOpen On-line\u201d hyperlink on the high proper. Suggestions welcome!<\/p>\n

committee; <\/span>ethics; <\/span>analysis; <\/span>generative ai; <\/span>functions; <\/span>sensible ideas; <\/span>engineering; <\/span>massive image concepts; <\/span>enjoyable; <\/span>reader updates; <\/span>jobs
\n<\/span><\/p><\/blockquote>\n

The complete program for the RSS Worldwide Convention (Bournmouth 7-10 September 2026) is <\/span>now out<\/p>\n

“Whether or not you wish to immerse your self within the latest methodological innovation, hear about actual world influence in well being, public coverage, and training, or be taught new expertise through our skilled improvement workshops, there may be loads to select from.”<\/p>\n

Our new journal is up and working <\/span>RSS: Knowledge Science and Synthetic Intelligence<\/p>\n

The second in our collection of important expertise for information scientists- “How one can write code you could be pleased with” – went forward on June sixteenth and was very profitable, with robust attendance and dialogue. Write as much as comply with.<\/p>\n

And the write up of the primary within the series- “How one can keep updated as an information scientist” – is now obtainable at <\/span>RealWorldDataScience<\/p>\n

A number of thrilling information science and AI occurring, as at all times!<\/p>\n

The Mythos saga- the US authorities will get concerned<\/p>\n

Anthropic’s new Mythos mannequin was launched.. after which the US authorities obtained anxious<\/p>\n

Implications for cyber safety: <\/span>Anthropic Embeds Engineers within the NSA to Deploy Mythos<\/p>\n

\n

Anthropic’s crimson staff mentioned Mythos Preview may discover and exploit zero-day vulnerabilities in ‘each main working system and each main net browser’ when a person directed it to. … A whole exploit pipeline towards a fancy Linux goal ran beneath a day at a price beneath $2,000, the put up mentioned, and roughly a thousand OpenBSD vulnerability searches value beneath $20,000 in complete.<\/p>\n<\/blockquote>\n

And what does this all imply for Europe? How essential is technical sovereignty and datacentre capability: <\/span>A viral doomsday situation goals to shake Europe out of its AI complacency<\/p>\n

\n

In a matter of years, America monopolises 70% of the world\u2019s “compute” \u2013 the semiconductor chips that fill the datacentres that energy AI fashions. Europe\u2019s financial system is in the meantime gasping for air, largely as a result of its corporations haven’t adopted AI.<\/p>\n<\/blockquote>\n

Actual-world misuse and the gradual erosion of public belief<\/p>\n

A essential turning level for skilled and legal accountability?: <\/span>Derbyshire police officer investigated over AI-generated \u2018evidential materials\u2019<\/p>\n

\n

A legal investigation has been launched into an allegation of perverting the course of justice after the alleged use of AI programs by an officer to create evidential materials in quite a few instances.<\/p>\n<\/blockquote>\n

Unintended penalties of changing human help with autonomous brokers: <\/span>Instagram is alerting customers who have been focused by hackers throughout AI chatbot assaults | TechCrunch<\/p>\n

\n

Hackers merely informed Meta\u2019s AI chatbot that they have been the house owners of the goal\u2019s account, and requested the bot to hyperlink that individual\u2019s account to an e-mail they managed. The chatbot complied with the request, permitting the hacker to reset the goal account\u2019s password and take management of the account.<\/p>\n<\/blockquote>\n

We’ve got reached a milestone the place AI reliably outperforms world-class human debaters: <\/span>AI programs out-persuade skilled people<\/p>\n

\n

We discovered that AI programs have been reliably extra persuasive than skilled people, even when skilled people selected their points, researched prematurely, underwent hours of stay, structured observe, and have been incentivised with \u00a31,000 money bonuses. … AI was practically 3x more practical than skilled canvassers from a UK fundraising agency at elevating real-money donations.<\/p>\n<\/blockquote>\n

How are functions altering our behaviour?<\/p>\n

What are the legalities of brokers searching websites on our behalf?: <\/span>Notes on Amazon v. Perplexity<\/p>\n

\n

The important thing perception right here is that the AI agent can course of the Website straight, talk straight with the person to find out the person’s intentions, after which work together with the Website utilizing the identical affordances as the positioning supplies for the person. The positioning does not want to show an API as a result of the Net interface turns into the API, and the mannequin supplies the UI.<\/p>\n<\/blockquote>\n

Does extra code essentially equal extra progress?: <\/span>The present influence of AI on engineering velocity<\/p>\n

\n

PR throughput is rising by about 10 to fifteen p.c. Throughout DX\u2019s pattern, most organizations noticed measurable enhancements, however the positive aspects have been far smaller than the 10x productiveness will increase typically cited in business headlines. The median enchancment was nearer to eight p.c. Whereas some organizations noticed bigger positive aspects, the everyday influence was extra incremental than transformational.<\/p>\n<\/blockquote>\n

Do you want an actual property agent to promote your own home?: <\/span>I Tried to Promote My Home With A.I.<\/p>\n

\n

It was all an indication of A.I.\u2019s potential to impart not simply info \u2014 the uncooked info it has ingested from Wikipedia or coaching supplies \u2014 however one thing nearer to knowledge. It granted me entry to a mind-set, a posture for navigating a nuanced and high-stakes interplay, that might usually be reserved for folks with innate expertise or a lot of earned expertise.<\/p>\n<\/blockquote>\n

What occurs subsequent?<\/p>\n

When AI begins autonomously designing its successors, the bottleneck for progress shifts solely from human ingenuity to compute energy.: <\/span>When AI builds itself Anthropic<\/p>\n

\n

For many of AI\u2019s historical past, people drove each step in its improvement cycle. However at Anthropic, we’re delegating a rising share of AI improvement to AI programs themselves, which is rushing up our work. Taken far sufficient, and given sufficient compute, that pattern factors to an AI system able to totally autonomously designing and creating its personal successor. That is referred to as recursive self-improvement.<\/p>\n<\/blockquote>\n

Can autonomous AI-powered killer drones take morality onboard?<\/p>\n

\n

Morality is deeply advanced, contested, culturally formed, and one thing most people by no means totally resolve, even for themselves. Maybe the actual query is whether or not we perceive morality nicely sufficient to codify it. Till we do, we can not count on machines to embody one thing we ourselves can not clearly articulate.<\/p>\n<\/blockquote>\n

Bettering reasoning<\/p>\n

Leverage human steerage to enhance RL exploration on arduous issues: <\/span>POPE: Studying to Motive on Arduous Issues through Privileged On-Coverage…<\/p>\n

\n

We present that mixing simple and arduous issues throughout RL coaching is counterproductive because of ray interference, the place optimization focuses on already-solvable issues in a method that actively inhibits progress on more durable ones. To handle this problem, we introduce Privileged On-Coverage Exploration (POPE), an strategy that leverages human- or different oracle options as privileged info to information exploration on arduous issues.<\/p>\n<\/blockquote>\n

Do fashions have to dream? Possibly they do!: <\/span>Language Fashions Want Sleep: Studying to Self-Modify and…<\/p>\n

\n

Impressed by human studying course of, we introduce a ”Sleep” paradigm that permits the fashions to repeatedly be taught, distill their short-term fragile recollections into secure long-term information with replay, and recursively enhance themselves with ”Dreaming” course of.<\/p>\n<\/blockquote>\n

Common analysis updates<\/p>\n

Diversifying sub-agents appears to assist with robustness: <\/span>Superhuman Protected and Agile Racing by way of Multi-Agent Reinforcement Studying<\/p>\n

\n

By league-based self-play, brokers evolve subtle anticipatory behaviors, together with proactive collision avoidance, overtaking, and dealing with multi-agent bodily interactions, together with aerodynamic downwash. Our brokers outperform a champion-level human pilot in multi-player races at speeds exceeding 22 m\/s, whereas concurrently lowering collision charges by 50 % in comparison with state-of-the-art single-agent baselines.<\/p>\n<\/blockquote>\n

Wonderful new picture corpus for permission authorized coaching: <\/span>GPIC: A Large Permissive Picture Corpus for Visible Era<\/p>\n

\n

GPIC includes various web photos captioned by a state-of-the-art vision-language mannequin, together with 100M coaching, 200K validation, and 1M check examples. Furthermore, all GPIC photos are permissively licensed for each analysis and business use. GPIC is safety-filtered, deduplicated, and centrally hosted on Hugging Face.<\/p>\n<\/blockquote>\n

A sobering reminder that automating security analysis would possibly simply disguise catastrophic flaws behind very convincing AI arguments.: <\/span>Automated alignment is more durable than you suppose<\/p>\n

\n

alignment analysis entails many hard-to-supervise fuzzy duties (duties with out clear analysis standards, for which human judgement is systematically flawed). Consequently, analysis outputs will include systematic, undetected errors, and even right outputs may very well be incorrectly aggregated into overconfident security assessments.<\/p>\n<\/blockquote>\n

Anthropic: Agentic Reliability and Regulatory Turbulence<\/p>\n

Fable, Mythos and all that: <\/span>Claude Fable 5 and Claude Mythos 5 Anthropic<\/p>\n

\n

Claude Fable 5’s reasoning is a transparent step past Opus 4.8. It really works at senior analysis scientist grade \u2014 selecting instructions, allocating assets, killing its incorrect beliefs, and producing novel first-principles outputs.<\/p>\n<\/blockquote>\n

Claude Fable autonomously engineering its personal debugging infrastructure.: <\/span>Claude Fable is relentlessly proactive<\/p>\n

\n
\n
\"Claude<\/div>\n<\/figure>\n<\/div>\n

The preliminary silent functionality throttling precipitated fairly a stir:<\/p>\n

Prime safety consultants are calling for the top of export controls to assist defenders match adversarial pace.: <\/span>Open Letter on Clear AI Cyber Protections<\/p>\n

\n

To tug one of the best capabilities away from defenders and not using a good motive when our adversaries are quickly advancing is harmful. … The justification for this unprecedented motion was that Fable supplies a singular \u201cuplift\u201d of capabilities past different AI fashions, however AI has been discovering bugs and producing working exploits at superhuman ranges since final 12 months.<\/p>\n<\/blockquote>\n

And many different updates and enhancements for Claud<\/p>\n

Google & Apple: Native Multimodality and Inference Pace<\/p>\n

Microsoft, Meta & OpenAI: Strategic Transformations<\/p>\n

OpenAI is evolving Codex right into a common workspace engine, democratising app creation for non-technical roles.: <\/span>Codex for each position, device, and workflow<\/p>\n

\n

Non-developers\u2014together with analysts, entrepreneurs, operators, designers, researchers, buyers, and bankers\u2014make up about 20% of total Codex customers\u2060 and are rising greater than 3x as quick as builders.<\/p>\n<\/blockquote>\n

Microsoft is transferring towards ‘Frontier Tuning,’ letting enterprises bake institutional information straight into mannequin weights.: <\/span>Constructing a hill-climbing machine: Launching seven new MAI fashions | Microsoft AI<\/p>\n

\n
\n
\"Building<\/div>\n<\/figure>\n<\/div>\n

Meta\u2019s pivot to the secretive TBD Lab indicators a tactical retreat from its conventional open-source advocacy.: <\/span>Inside Meta’s makes an attempt to play catch-up with AI – Ars Technica<\/p>\n

\n

The very small staff the place everyone seems to be \u2018cracked\u2019 is at all times going to maneuver sooner than the massive org the place duty is distributed.<\/p>\n<\/blockquote>\n

Chinese language Frontier Fashions: Open Weights and Sparse Architectures<\/p>\n

GLM-5.2 units a formidable new ceiling for open fashions, efficiently narrowing the hole to absolutely the frontier: <\/p>\n

GLM-5.2 Is The New Greatest Open Mannequin<\/p>\n

\n
\n
\"GLM-5.2<\/div>\n<\/figure>\n<\/div>\n

The IndexShare structure ensures reliability over time with a large 1-million-token context window.: <\/span>GLM-5.2: Constructed for Lengthy-Horizon Duties<\/p>\n

\n
\n
\"GLM-5.2:<\/div>\n<\/figure>\n<\/div>\n

It’s arguably essentially the most highly effective text-only open weights mannequin obtainable, although it stays notably token-hungry.: <\/span>GLM-5.2 might be essentially the most highly effective text-only open weights LLM<\/p>\n

\n

GLM-5.2 is the main open weights mannequin on the Intelligence Index v4.1. At 51, it leads MiniMax-M3 (44), DeepSeek V4 Professional (max, 44) and Kimi K2.6 (43). . . GLM-5.2 makes use of extra output tokens per activity than different main open weights fashions: the mannequin makes use of 43k output tokens per Intelligence Index activity, up from GLM-5.1 (26k).<\/p>\n<\/blockquote>\n

Moonshot\u2019s new MoE mannequin prioritises reasoning effectivity, reducing the token overhead for repository-scale refactoring.: <\/span>Moonshot AI Releases Kimi K2.7-Code: a Coding Mannequin Reporting +21.8% on Kimi Code Bench v2 Over K2.6 – MarkTechPost<\/p>\n

\n

K2.7-Code is a Combination-of-Consultants mannequin. It holds 1T complete parameters and prompts 32B per token. The design makes use of 384 consultants, with 8 chosen per token and 1 shared. It has 61 layers, together with 1 dense layer. Consideration makes use of MLA, and the feed-forward path makes use of SwiGLU. A MoonViT imaginative and prescient encoder provides 400M parameters for picture and video enter.<\/p>\n<\/blockquote>\n

MiniMax M3 packs a million-token context and native multimodality into an environment friendly open-weight package deal.: <\/span>MiniMax M3: Frontier Coding, 1M Context, Native Multimodality \u2014 All in One Mannequin – MiniMax Analysis | MiniMax<\/p>\n

\n
\n
\"MiniMax<\/div>\n<\/figure>\n<\/div>\n

Qwen3.7-Plus indicators a pivot towards autonomous company, enabling fashions to confirm and iterate on their very own code.: <\/span>Alibaba’s Qwen Crew Launches Qwen3.7-Plus, Including Imaginative and prescient, Deep Reasoning, Instrument Invocation, and Autonomous Iteration on the Bailian Platform – MarkTechPost<\/p>\n

\n

Self-programming means the mannequin writes and revises its personal code. Instrument invocation means it calls exterior features or APIs. Verification and testing means it runs outputs and checks outcomes. Autonomous iteration means it loops till the duty is completed. Collectively, they describe a mannequin constructed to behave, not simply reply.<\/p>\n<\/blockquote>\n

NVIDIA retains producing good fashions<\/p>\n

Cohere proves that environment friendly Combination-of-Consultants can ship enterprise-grade coding capabilities on modest, native {hardware}.: <\/span>North Mini Code: Agentic Coding Mannequin for Builders | Cohere<\/p>\n

\n
\n
\"North<\/div>\n<\/figure>\n<\/div>\n

Benchmarks, Ensembles & Specialised Instruments<\/p>\n

Cognition\u2019s new benchmark reveals that frontier fashions nonetheless wrestle with the ‘mergeability’ required for manufacturing codebases.: <\/span>Introducing FrontierCode | Cognition<\/p>\n

\n

What units FrontierCode aside is the eye to element. Every activity is calibrated to a depth that merely hasn\u2019t been seen earlier than in LLM benchmarking. We ought to be transferring away from benchmarks that may be gamed and as an alternative utilizing ones like FrontierCode to reveal real mannequin intelligence and creativity.<\/p>\n<\/blockquote>\n

The following efficiency leap would possibly come from fusing various mannequin panels fairly than scaling particular person architectures.: <\/span>Surpassing Frontier Efficiency with Fusion \u2014 OpenRouter Weblog<\/p>\n

\n

Fable 5 + GPT-5.5 fused collectively scored 69.0%, surpassing each particular person mannequin, together with Fable 5 alone at 65.3%. A funds panel (Gemini 3 Flash, Kimi K2.6, and DeepSeek V4 Professional) beat GPT-5.5 and Opus 4.8. It got here inside 1% of Fable 5\u2019s rating whereas being 50% of the price.<\/p>\n<\/blockquote>\n

This tiny 3B mannequin is inflicting all kinds of debate.: <\/span>Why Weibo\u2019s tiny VibeThinker-3B has the AI world arguing over benchmarks once more<\/p>\n

\n

The event of compact fashions is now not merely a passive compromise for deployment effectivity or value management; it emerges as a promising analysis trajectory that’s essentially complementary to the normal parameter scaling paradigm.<\/p>\n<\/blockquote>\n

From proteins to planetary security: Utilizing AI to map and engineer our bodily actuality<\/p>\n

New open-source protein folding fashions: <\/span>Biohub releases a world mannequin of protein biology<\/p>\n

\n
\n
\"Biohub<\/div>\n<\/figure>\n<\/div>\n

Designing ‘super-antigens’ from scratch: <\/span>‘World-first’ vaccine designed by synthetic intelligence<\/p>\n

\n

These genetic codes have been analysed by a synthetic intelligence. It then designed a ‘super-antigen’ that might prepare the immune system in such a method it gave safety towards the entire household of viruses \u2013 even when they mutated or a brand new an infection jumped from animals to folks.<\/p>\n<\/blockquote>\n

Serving to physicists simulate the advanced plasma round occasion horizons.: <\/span>How an astrophysicist makes use of Codex to assist simulate black holes<\/p>\n

\n

Codex generated many potential approaches\u2014not all of them right. \u2018However that\u2019s okay,\u2019 Chan mentioned. \u2018Most scientific concepts fail. What issues is that these algorithms are testable. As soon as you discover one which works, it will probably doubtlessly unlock simulations that have been beforehand unimaginable.\u2019<\/p>\n<\/blockquote>\n

So impressive- turning ubiquitous shopper {hardware} right into a life-saving distributed seismic community.: <\/span>How Telephones Alerted Hundreds of thousands Earlier than Quakes Shook Venezuela<\/p>\n

\n

Venezuela doesn’t have a nationwide early warning system of its personal, however folks with Android telephones obtained alerts from Google\u2019s Earthquake Alerts system, which might pull information from greater than two billion telephones geared up with built-in accelerometers. The identical sensor that detects rotation on the display can even sense vibrations from seismic waves.<\/p>\n<\/blockquote>\n

A foundational layer for international meals safety.: <\/span>Agricultural Discipline Boundaries<\/p>\n

\n
\n
\"Agricultural<\/div>\n<\/figure>\n<\/div>\n

A unbelievable use of real-time laptop imaginative and prescient to steadiness industrial maritime logistics with pressing ecological preservation.: <\/span>AI-powered whale-spotting tech might assist save San Francisco Bay\u2019s grey whales<\/p>\n

\n

The brand new expertise combines round the clock thermal cameras deployed at totally different areas within the bay with AI to detect whales that could be so far as 7 kilometers away. As soon as the whale detection is confirmed by scientists, an alert goes out to warn vessels within the space to decelerate or change course to keep away from a collision.<\/p>\n<\/blockquote>\n

Generalists vs specialists<\/p>\n

Sakana reveals that sensible ensembles can beat frontier monoliths.: <\/span>Sakana Fugu \u2014 Multi-agent System as A Mannequin<\/p>\n

\n

Sakana Fugu achieves superior efficiency by dynamically coordinating and orchestrating a various pool of highly effective fashions. As a substitute of utilizing area information to prescribe staff group, roles, or workflows, Fugu learns to dynamically assemble brokers from a pool and coordinate them by way of non-obvious however extremely environment friendly collaboration patterns.<\/p>\n<\/blockquote>\n

A rigorous new examine means that general-purpose mannequin scale is now extra essential for medical competency than specialised tuning.: <\/span>Common-purpose giant language fashions outperform specialised scientific AI instruments on medical benchmarks<\/p>\n

\n
\n
\"General-purpose<\/div>\n<\/figure>\n<\/div>\n

Anthropic has proven that Claude can outperform specialist software program in NMR spectral prediction.: <\/span>Making Claude a chemist Anthropic<\/p>\n

\n

In the end, we discovered that for routine information prediction Opus 4.7\u2014a general-purpose mannequin with out chemistry-specific fine-tuning\u2014is now pretty much as good as or higher than ChemDraw and MestReNova on common. Moreover, Claude can even work the issue in reverse, proposing a construction from NMR information alone.<\/p>\n<\/blockquote>\n

Past the search bar \u2013 utilizing AI as a classy reasoning sandbox {and professional} critic<\/p>\n

Touchdown a task at a high AI lab requires treating the interview itself as a definite engineering challenge.: <\/span>ML Job Interviews: The Final Information \u2013 Silvia Sapora<\/p>\n

\n

I do know extraordinarily spectacular researchers who have been rejected in interviews just because they did not put together. Working with ML day in and day trip is just not the identical as being able to implement consideration from scratch, derive the backward go, or code flash consideration. Allocate at the very least a month of normal examine time.<\/p>\n<\/blockquote>\n

There is no such thing as a longer an excuse for sloppy analysis manuscripts when LLMs act as world-class technical critics.: <\/span>Why Teachers Ought to Use AI for Writing: A Case Examine<\/p>\n

\n

The standard and amount of ChatGPT\u2019s suggestions left me surprised. I had no concept that I may make so many blatant errors. This sobering expertise leads me to consider that educational authors can typically enhance their manuscripts tremendously by asking an LLM for paragraph-by-paragraph suggestions.<\/p>\n<\/blockquote>\n

Critiquing survey methodology.: <\/span>AI and Survey Sampling Issues \u2014 Sharon Lohr<\/p>\n

\n

College students have to know find out how to choose, analyze, and critique samples themselves to have the ability to distinguish Gemini\u2019s useful solutions from the nonsense. However Gemini may assist develop college students\u2019 understanding of the topic throughout in-class actions. It will possibly discover examples of survey datasets with varied traits, can write code for graphing or analyzing survey information, and its critiques can spur class dialogue.<\/p>\n<\/blockquote>\n

Architecting the agentic loop \u2013 constructing strong harnesses<\/p>\n

Andrew Ng’s tackle an open harness- value experimenting with: <\/span>OpenCoworker<\/p>\n

\n

OpenCoworker is a desktop AI agent that may not solely chat, but in addition do deep analysis and perform duties for you in your laptop. It will possibly learn information (with permission) to realize context, learn\/ship messages (slack, e-mail, and many others.), and create actual deliverables like PDF studies, paperwork, spreadsheets. It additionally helps scheduled automations, reminiscent of offering you a each day information abstract.<\/p>\n<\/blockquote>\n

Smaller, native fashions can outmatch frontier giants when you prioritise structured information structure over uncooked mannequin scale.: <\/span>Information Brokers: Beat Frontier Fashions with Higher Construction<\/p>\n

\n
\n
\"Knowledge<\/div>\n<\/figure>\n<\/div>\n

Bridging the hole between semantic reasoning and real-time bodily motion: <\/span>An Overview of Fashionable AI Robotics from First Ideas | Interlatent Weblog<\/p>\n

\n

Whereas a language mannequin can take three seconds to consider its subsequent token and nobody is harmed, a robotic pouring espresso can not. The cup is already transferring and the actions should be generated mid-event. The operate does not simply should be right, it needs to be quick sufficient that correctness nonetheless issues by the point the reply arrives.<\/p>\n<\/blockquote>\n

Perceive the internal mechanics of state-of-the-art diffusion transformers by way of this clear implementation.: <\/span>purohit10saurabh\/minFLUX: A hackable implementation of FLUX diffusion fashions<\/p>\n

\n

A simplified academic PyTorch implementation of FLUX.1 and FLUX.2 diffusion transformers (DiT) by Black Forest Labs. Constructed for understanding rectified circulate matching, joint consideration, and the important thing design selections behind FLUX with verifiable line-by-line supply mappings to the official codebases.<\/p>\n<\/blockquote>\n

Statistically sound foundations<\/p>\n

Classes from the entrance line (from our very personal Martin Goodson): <\/span>Growing a excessive efficiency machine studying algorithm<\/p>\n

\n

Sadly, removed from being easy, the analysis challenge became a multitude of complexity. At each flip, our technical strategy gave the impression to be the improper one. Greater than as soon as, team-members informed me that what we have been attempting to do was unimaginable. However the journey was wealthy with hard-won classes, which I need to share with you right here.<\/p>\n<\/blockquote>\n

Understanding Meta’s JEPA by way of 90-year-old foundations in classical canonical correlation evaluation.: <\/span>The 90-year-old thought behind JEPA fashions: Canonical Correlation Evaluation (CCA) \u2013 Shon Czinner\u2019s Weblog<\/p>\n

\n

In the end, these fashions all have the identical goal operate launched by CCA: discover the transformations that end in maximal correlation between units of multidimensional information.<\/p>\n<\/blockquote>\n

Detecting mannequin drift turns into considerably extra strong when utilizing a symmetric and secure diagnostic like Jensen-Shannon divergence.: <\/span>Jensen-Shannon Divergence | Josh Lospinoso<\/p>\n

\n

Jensen-Shannon divergence is helpful when two discrete or binned distributions want a bounded, symmetric discrepancy rating. It seems in prepare\/check drift, subject or class combine monitoring, histogram comparability, token checks, and cohort composition overview. If ignored, two teams can have comparable averages whereas allocating likelihood mass to totally different bins.<\/p>\n<\/blockquote>\n

Deal with imprecise spatial measurements by treating coordinates as latent variables inside a rigorous Bayesian Gaussian Course of framework.: <\/span>Do not know the place your information is from? Bayesian modeling for unknown coordinates | Christopher Krapu<\/p>\n

\n

Spatial location error adjustments the covariance and prediction downside itself… Bayesian modeling with applicable priors lets us modify and alter practically any a part of the mannequin, on condition that we now have some thought of find out how to signify our assumptions as a part of the mannequin course of. Then, we use Monte Carlo strategies to show the inference crank and acquire dependable parameter estimates.<\/p>\n<\/blockquote>\n

Refining the machine <\/p>\n

Discovering optimum tokenisation approaches with linear programming: <\/span>Discovering Optimum Tokenizers<\/p>\n

\n

On this formulation, there is a \u201ccoloration\u201d variable for every doable vocabulary entry. Specifically, we create one coloration variable for each distinctive substring of the dataset. A coloration variable is 1 if the corresponding byte sequence is within the vocabulary, or 0 in any other case. We add a single constraint to power the sum of coloration variables to equal the goal vocabulary measurement.<\/p>\n<\/blockquote>\n

A helpful primer on post-training reasoning information: <\/span>A Primer in Put up-Coaching Reasoning Knowledge: What We Know About How It Works<\/p>\n

\n

Put up-training has change into a main driver of current progress in giant reasoning fashions, and reasoning information are sometimes the important thing variable figuring out whether or not this stage succeeds. Work on post-training reasoning information has grown quickly, but this literature stays scattered throughout dataset papers, reinforcement-learning recipes, reward-model research, benchmarks, and frontier system studies.<\/p>\n<\/blockquote>\n

Fable\/Mythos the sport changer?<\/p>\n

How Sam (OpenAI) and Dario (Anthropic) see the world<\/p>\n

Constructed to profit everybody: our plan – Sam Altman and Jakub Pachocki<\/span><\/p>\n

\n

We consider that AI doing AI analysis will change into the figuring out issue of the tempo of progress inside the subsequent few years. That issues as a result of alignment is itself a tough analysis downside. To make quick and deep progress, our researchers will want AI programs that may assist check concepts, discover errors, discover alternate options, and iterate alongside us.<\/p>\n<\/blockquote>\n

Dario Amodei \u2014 Coverage on the AI Exponential – Dario Amodei<\/span><\/p>\n

\n

Frontier AI fashions, like airplanes, ought to be required to undergo technical testing and auditing, and their launch ought to be blocked or reversed as a menace to public security if they don’t meet excessive requirements of security. I’m grateful to see the Trump administration\u2019s Government Order transfer incrementally in direction of a larger position for presidency in AI, although Anthropic\u2019s proposal recommends even additional motion.<\/p>\n<\/blockquote>\n

Meals for thought<\/p>\n

The state of the AI financial system – Azeem Azhar, William Gildea, and Nathan Warren<\/span><\/p>\n

\n
\n
\"\ud83d\udd2e<\/div>\n<\/figure>\n<\/div>\n

AI for Bio has a Fuzzy API downside – Ankit Gupta<\/span><\/p>\n

\n

The output of goal discovery is just not actually a goal. It’s a probabilistic speculation that modulating some organic course of, in some route, in some tissue, in some affected person subset, at some illness stage, will produce a helpful scientific impact with out unacceptable toxicity.<\/p>\n<\/blockquote>\n

Open and closed fashions are on totally different exponentials – Nathan Lambert<\/span><\/p>\n

\n

The closed fashions hit unimaginable product-market match with the present brokers, beginning their built-in exponential by monetizing the highest finish of the information work. The open mannequin financial system will take far longer, however it’ll even be way more satisfying to comply with, because it tracks the broader diffusion of AI into your complete financial system and world.<\/p>\n<\/blockquote>\n

The Structural Boundaries to AI Attorneys – Sean A. Harrington<\/span><\/p>\n

\n

Structural obstacles make authorized observe proof against technological diffusion in ways in which different industries don\u2019t face. Understanding these obstacles issues as a result of regulation is the place AI meets civic infrastructure. Courts, contracts, laws, and rights circulate by way of attorneys. If AI can\u2019t diffuse by way of regulation, its broader social influence will stay constrained.<\/p>\n<\/blockquote>\n

Why Does Everybody Hate AI? – Paul Krugman<\/span><\/p>\n

\n
\n
\"Why<\/div>\n<\/figure>\n<\/div>\n

Co-Existence and the Finish of Co-Intelligence – Ethan Mollick<\/span><\/p>\n

\n
\n
\"Co-Existence<\/div>\n<\/figure>\n<\/div>\n

Opinion | What 370,000 Faculty Essays Inform Us About A.I.\u2019s Results on Creativity – Rebecca Winthrop<\/span><\/p>\n

\n

A.I.\u2019s easy sentences, elegant transitions and wealthy vocabulary give the phantasm of expansive creativity and individuality. However the underlying concepts typically converge into just a few homogenized classes. … In a separate examine, the staff discovered that human-written essays supplied as much as eight instances extra new concepts than these produced by A.I.<\/p>\n<\/blockquote>\n

Enjoyable initiatives<\/p>\n

Ever questioned how an AI would discuss if it solely ever learn Dickens?: <\/span>Making a classic LLM from scratch;<\/p>\n

\n

I did not need my LLM to find out about computer systems, atomic bombs and space-ships, so I had no alternative however to make my very own… From the beginning, I knew that the information is a very powerful: rubbish in -> rubbish out. I did heaps and plenty of experiments and iterations, I performed with DBs like Qdrant, Zvec, Lance, ValKey, and LevelDB for storing the datasets.<\/p>\n<\/blockquote>\n

Wonderful interactive exploration that breaks down the Fourier evaluation behind your favorite digital sounds.: <\/span>How The Heck Do Synthesizers Work? (An Interactive Exploration)<\/p>\n

\n

It is a consequence of a theorem confirmed by Joseph Fourier within the early 1800s: any periodic waveform could be decomposed right into a sum of sine waves at totally different frequencies, amplitudes, and phases.<\/p>\n<\/blockquote>\n

Intelligent Raspberry Pi challenge makes use of edge AI to show yard fowl calls into stunning Japanese-style art work.: <\/span>Avian Guests – Teddy Warner<\/p>\n

\n

The packing algorithm itself is a center-out spiral: tiles get sorted by space descending, the biggest is positioned on the heart of mass, and every subsequent tile spirals outward from the middle till discovering a place the place its masks doesn\u2019t intersect any already-placed masks.<\/p>\n<\/blockquote>\n

Prime-Minister turnover by way of the ages: <\/span>United Kingdom prime ministers<\/p>\n

\n

The truth is, within the late twentieth century with Thatcher and Blair, the UK confronted a interval of unusually sluggish turnover of prime ministers. However that was a formative interval within the lifetime of lots of as we speak\u2019s political commentators, so its not stunning that the present fast turnover comes throughout as a shock.<\/p>\n<\/blockquote>\n

So that you need to construct a robotic?.: <\/span>A Newbie’s Information to Robotics {Hardware} | Interlatent Weblog<\/p>\n

\n

The command that’s despatched is just not the movement or the torque, however end-state outcomes. An instruction to maneuver a joint to a given place passes by way of friction and cargo earlier than it turns into actual motion. This hole between commanded and achieved movement is strictly why learning-based controllers like Motion Chunking with Transformers (ACT) predict quick sequences of actions fairly than one step at a time to maintain small errors from compounding over an extended horizon.<\/p>\n<\/blockquote>\n

Eran Raviv, Knowledgeable researcher (DS) at APG Asset Administration, has revealed an fascinating pape on covariance estimation for vast data- see <\/span>right here<\/p>\n

The Job market is a bit quiet – tell us when you’ve got any openings you\u2019d wish to promote<\/p>\n

Once more, hope you discovered this convenient. Please do ship on to your friends- we wish to construct a robust group of knowledge science practitioners- and join future updates right here<\/p>\n

The views expressed are our personal and don’t essentially signify these of the RSS<\/p>\n<\/div>\n


\n
Source link <\/a><\/p>\n","protected":false},"excerpt":{"rendered":"

Hello everyone- Fairly a month! World Cup kicking off, offers (or no offers) within the Straits of Hormuz, a possible new PM, and the most well liked days on file… positively time for a quiet learn within the shade with all of the goings on within the vast world of AI and Knowledge Science. A […]<\/p>\n","protected":false},"author":1,"featured_media":1942,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"fifu_image_url":"https:\/\/substackcdn.com\/image\/fetch\/$s_!6uEr!,w_1200,h_675,c_fill,f_jpg,q_auto:good,fl_progressive:steep,g_auto\/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F67ad08f8-c41a-4fe4-93fb-072143bc28d7_1604x794.png","fifu_image_alt":"","jnews-multi-image_gallery":[],"jnews_single_post":[],"jnews_primary_category":[],"jnews_override_bookmark_settings":[],"jnews_social_meta":[],"jnews_override_counter":[],"footnotes":""},"categories":[4],"tags":[2438,212,213,214],"class_list":["post-1940","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-ethics-policy","tag-july","tag-newsletter","tag-piers","tag-stobbs"],"yoast_head":"\nJuly 2026 Publication - by Piers Stobbs - Future News 24<\/title>\n<meta name=\"robots\" content=\"index, follow, max-snippet:-1, max-image-preview:large, max-video-preview:-1\" \/>\n<link rel=\"canonical\" href=\"https:\/\/futurenews24.com\/index.php\/2026\/07\/01\/july-2026-newsletter\/\" \/>\n<meta property=\"og:locale\" content=\"en_US\" \/>\n<meta property=\"og:type\" content=\"article\" \/>\n<meta property=\"og:title\" content=\"July 2026 Publication - by Piers Stobbs - Future News 24\" \/>\n<meta property=\"og:description\" content=\"Hello everyone- Fairly a month! World Cup kicking off, offers (or no offers) within the Straits of Hormuz, a possible new PM, and the most well liked days on file… positively time for a quiet learn within the shade with all of the goings on within the vast world of AI and Knowledge Science. A […]\" \/>\n<meta property=\"og:url\" content=\"https:\/\/futurenews24.com\/index.php\/2026\/07\/01\/july-2026-newsletter\/\" \/>\n<meta property=\"og:site_name\" content=\"Future News 24\" \/>\n<meta property=\"article:published_time\" content=\"2026-07-01T05:01:00+00:00\" \/>\n<meta property=\"article:modified_time\" content=\"2026-07-06T06:59:09+00:00\" \/>\n<meta property=\"og:image\" content=\"https:\/\/substackcdn.com\/image\/fetch\/$s_!6uEr!,w_1200,h_675,c_fill,f_jpg,q_auto:good,fl_progressive:steep,g_auto\/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F67ad08f8-c41a-4fe4-93fb-072143bc28d7_1604x794.png\" \/><meta property=\"og:image\" content=\"https:\/\/substackcdn.com\/image\/fetch\/$s_!6uEr!,w_1200,h_675,c_fill,f_jpg,q_auto:good,fl_progressive:steep,g_auto\/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F67ad08f8-c41a-4fe4-93fb-072143bc28d7_1604x794.png\" \/>\n<meta name=\"author\" content=\"Future News 24\" \/>\n<meta name=\"twitter:card\" content=\"summary_large_image\" \/>\n<meta name=\"twitter:image\" content=\"https:\/\/substackcdn.com\/image\/fetch\/$s_!6uEr!,w_1200,h_675,c_fill,f_jpg,q_auto:good,fl_progressive:steep,g_auto\/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F67ad08f8-c41a-4fe4-93fb-072143bc28d7_1604x794.png\" \/>\n<meta name=\"twitter:label1\" content=\"Written by\" \/>\n\t<meta name=\"twitter:data1\" content=\"Future News 24\" \/>\n\t<meta name=\"twitter:label2\" content=\"Est. reading time\" \/>\n\t<meta name=\"twitter:data2\" content=\"23 minutes\" \/>\n<script type=\"application\/ld+json\" class=\"yoast-schema-graph\">{\"@context\":\"https:\\\/\\\/schema.org\",\"@graph\":[{\"@type\":\"Article\",\"@id\":\"https:\\\/\\\/futurenews24.com\\\/index.php\\\/2026\\\/07\\\/01\\\/july-2026-newsletter\\\/#article\",\"isPartOf\":{\"@id\":\"https:\\\/\\\/futurenews24.com\\\/index.php\\\/2026\\\/07\\\/01\\\/july-2026-newsletter\\\/\"},\"author\":{\"name\":\"Future News 24\",\"@id\":\"https:\\\/\\\/futurenews24.com\\\/#\\\/schema\\\/person\\\/cecad1bde21cfc357cf70128144d6c83\"},\"headline\":\"July 2026 Publication – by Piers Stobbs\",\"datePublished\":\"2026-07-01T05:01:00+00:00\",\"dateModified\":\"2026-07-06T06:59:09+00:00\",\"mainEntityOfPage\":{\"@id\":\"https:\\\/\\\/futurenews24.com\\\/index.php\\\/2026\\\/07\\\/01\\\/july-2026-newsletter\\\/\"},\"wordCount\":4780,\"commentCount\":0,\"publisher\":{\"@id\":\"https:\\\/\\\/futurenews24.com\\\/#organization\"},\"image\":{\"@id\":\"https:\\\/\\\/futurenews24.com\\\/index.php\\\/2026\\\/07\\\/01\\\/july-2026-newsletter\\\/#primaryimage\"},\"thumbnailUrl\":\"https:\\\/\\\/substackcdn.com\\\/image\\\/fetch\\\/$s_!6uEr!,w_1200,h_675,c_fill,f_jpg,q_auto:good,fl_progressive:steep,g_auto\\\/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F67ad08f8-c41a-4fe4-93fb-072143bc28d7_1604x794.png\",\"keywords\":[\"July\",\"Newsletter\",\"Piers\",\"Stobbs\"],\"articleSection\":[\"Ethics & Policy\"],\"inLanguage\":\"en-US\",\"potentialAction\":[{\"@type\":\"CommentAction\",\"name\":\"Comment\",\"target\":[\"https:\\\/\\\/futurenews24.com\\\/index.php\\\/2026\\\/07\\\/01\\\/july-2026-newsletter\\\/#respond\"]}]},{\"@type\":\"WebPage\",\"@id\":\"https:\\\/\\\/futurenews24.com\\\/index.php\\\/2026\\\/07\\\/01\\\/july-2026-newsletter\\\/\",\"url\":\"https:\\\/\\\/futurenews24.com\\\/index.php\\\/2026\\\/07\\\/01\\\/july-2026-newsletter\\\/\",\"name\":\"July 2026 Publication - by Piers Stobbs - Future News 24\",\"isPartOf\":{\"@id\":\"https:\\\/\\\/futurenews24.com\\\/#website\"},\"primaryImageOfPage\":{\"@id\":\"https:\\\/\\\/futurenews24.com\\\/index.php\\\/2026\\\/07\\\/01\\\/july-2026-newsletter\\\/#primaryimage\"},\"image\":{\"@id\":\"https:\\\/\\\/futurenews24.com\\\/index.php\\\/2026\\\/07\\\/01\\\/july-2026-newsletter\\\/#primaryimage\"},\"thumbnailUrl\":\"https:\\\/\\\/substackcdn.com\\\/image\\\/fetch\\\/$s_!6uEr!,w_1200,h_675,c_fill,f_jpg,q_auto:good,fl_progressive:steep,g_auto\\\/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F67ad08f8-c41a-4fe4-93fb-072143bc28d7_1604x794.png\",\"datePublished\":\"2026-07-01T05:01:00+00:00\",\"dateModified\":\"2026-07-06T06:59:09+00:00\",\"breadcrumb\":{\"@id\":\"https:\\\/\\\/futurenews24.com\\\/index.php\\\/2026\\\/07\\\/01\\\/july-2026-newsletter\\\/#breadcrumb\"},\"inLanguage\":\"en-US\",\"potentialAction\":[{\"@type\":\"ReadAction\",\"target\":[\"https:\\\/\\\/futurenews24.com\\\/index.php\\\/2026\\\/07\\\/01\\\/july-2026-newsletter\\\/\"]}]},{\"@type\":\"ImageObject\",\"inLanguage\":\"en-US\",\"@id\":\"https:\\\/\\\/futurenews24.com\\\/index.php\\\/2026\\\/07\\\/01\\\/july-2026-newsletter\\\/#primaryimage\",\"url\":\"https:\\\/\\\/substackcdn.com\\\/image\\\/fetch\\\/$s_!6uEr!,w_1200,h_675,c_fill,f_jpg,q_auto:good,fl_progressive:steep,g_auto\\\/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F67ad08f8-c41a-4fe4-93fb-072143bc28d7_1604x794.png\",\"contentUrl\":\"https:\\\/\\\/substackcdn.com\\\/image\\\/fetch\\\/$s_!6uEr!,w_1200,h_675,c_fill,f_jpg,q_auto:good,fl_progressive:steep,g_auto\\\/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F67ad08f8-c41a-4fe4-93fb-072143bc28d7_1604x794.png\"},{\"@type\":\"BreadcrumbList\",\"@id\":\"https:\\\/\\\/futurenews24.com\\\/index.php\\\/2026\\\/07\\\/01\\\/july-2026-newsletter\\\/#breadcrumb\",\"itemListElement\":[{\"@type\":\"ListItem\",\"position\":1,\"name\":\"Home\",\"item\":\"https:\\\/\\\/futurenews24.com\\\/\"},{\"@type\":\"ListItem\",\"position\":2,\"name\":\"July 2026 Publication – by Piers Stobbs\"}]},{\"@type\":\"WebSite\",\"@id\":\"https:\\\/\\\/futurenews24.com\\\/#website\",\"url\":\"https:\\\/\\\/futurenews24.com\\\/\",\"name\":\"Future News 24\",\"description\":\"The Smart Hub for AI and Next-Gen Innovation\",\"publisher\":{\"@id\":\"https:\\\/\\\/futurenews24.com\\\/#organization\"},\"potentialAction\":[{\"@type\":\"SearchAction\",\"target\":{\"@type\":\"EntryPoint\",\"urlTemplate\":\"https:\\\/\\\/futurenews24.com\\\/?s={search_term_string}\"},\"query-input\":{\"@type\":\"PropertyValueSpecification\",\"valueRequired\":true,\"valueName\":\"search_term_string\"}}],\"inLanguage\":\"en-US\"},{\"@type\":\"Organization\",\"@id\":\"https:\\\/\\\/futurenews24.com\\\/#organization\",\"name\":\"Future News 24\",\"url\":\"https:\\\/\\\/futurenews24.com\\\/\",\"logo\":{\"@type\":\"ImageObject\",\"inLanguage\":\"en-US\",\"@id\":\"https:\\\/\\\/futurenews24.com\\\/#\\\/schema\\\/logo\\\/image\\\/\",\"url\":\"https:\\\/\\\/futurenews24.com\\\/wp-content\\\/uploads\\\/2026\\\/06\\\/fn24-favicon.png\",\"contentUrl\":\"https:\\\/\\\/futurenews24.com\\\/wp-content\\\/uploads\\\/2026\\\/06\\\/fn24-favicon.png\",\"width\":250,\"height\":250,\"caption\":\"Future News 24\"},\"image\":{\"@id\":\"https:\\\/\\\/futurenews24.com\\\/#\\\/schema\\\/logo\\\/image\\\/\"}},{\"@type\":\"Person\",\"@id\":\"https:\\\/\\\/futurenews24.com\\\/#\\\/schema\\\/person\\\/cecad1bde21cfc357cf70128144d6c83\",\"name\":\"Future News 24\",\"image\":{\"@type\":\"ImageObject\",\"inLanguage\":\"en-US\",\"@id\":\"https:\\\/\\\/secure.gravatar.com\\\/avatar\\\/d57f07142d73cb5503ab2446ea7bc9ef3d0a5ba378d64a6157692311e42bf097?s=96&d=mm&r=g\",\"url\":\"https:\\\/\\\/secure.gravatar.com\\\/avatar\\\/d57f07142d73cb5503ab2446ea7bc9ef3d0a5ba378d64a6157692311e42bf097?s=96&d=mm&r=g\",\"contentUrl\":\"https:\\\/\\\/secure.gravatar.com\\\/avatar\\\/d57f07142d73cb5503ab2446ea7bc9ef3d0a5ba378d64a6157692311e42bf097?s=96&d=mm&r=g\",\"caption\":\"Future News 24\"},\"sameAs\":[\"https:\\\/\\\/futurenews24.com\"],\"url\":\"https:\\\/\\\/futurenews24.com\\\/index.php\\\/author\\\/mridulpahuja20\\\/\"}]}<\/script>\n<!-- \/ Yoast SEO plugin. -->","yoast_head_json":{"title":"July 2026 Publication - by Piers Stobbs - Future News 24","robots":{"index":"index","follow":"follow","max-snippet":"max-snippet:-1","max-image-preview":"max-image-preview:large","max-video-preview":"max-video-preview:-1"},"canonical":"https:\/\/futurenews24.com\/index.php\/2026\/07\/01\/july-2026-newsletter\/","og_locale":"en_US","og_type":"article","og_title":"July 2026 Publication - by Piers Stobbs - Future News 24","og_description":"Hello everyone- Fairly a month! World Cup kicking off, offers (or no offers) within the Straits of Hormuz, a possible new PM, and the most well liked days on file… positively time for a quiet learn within the shade with all of the goings on within the vast world of AI and Knowledge Science. A […]","og_url":"https:\/\/futurenews24.com\/index.php\/2026\/07\/01\/july-2026-newsletter\/","og_site_name":"Future News 24","article_published_time":"2026-07-01T05:01:00+00:00","article_modified_time":"2026-07-06T06:59:09+00:00","og_image":[{"url":"https:\/\/substackcdn.com\/image\/fetch\/$s_!6uEr!,w_1200,h_675,c_fill,f_jpg,q_auto:good,fl_progressive:steep,g_auto\/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F67ad08f8-c41a-4fe4-93fb-072143bc28d7_1604x794.png","type":"","width":"","height":""},{"url":"https:\/\/substackcdn.com\/image\/fetch\/$s_!6uEr!,w_1200,h_675,c_fill,f_jpg,q_auto:good,fl_progressive:steep,g_auto\/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F67ad08f8-c41a-4fe4-93fb-072143bc28d7_1604x794.png","type":"","width":"","height":""}],"author":"Future News 24","twitter_card":"summary_large_image","twitter_image":"https:\/\/substackcdn.com\/image\/fetch\/$s_!6uEr!,w_1200,h_675,c_fill,f_jpg,q_auto:good,fl_progressive:steep,g_auto\/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F67ad08f8-c41a-4fe4-93fb-072143bc28d7_1604x794.png","twitter_misc":{"Written by":"Future News 24","Est. reading time":"23 minutes"},"schema":{"@context":"https:\/\/schema.org","@graph":[{"@type":"Article","@id":"https:\/\/futurenews24.com\/index.php\/2026\/07\/01\/july-2026-newsletter\/#article","isPartOf":{"@id":"https:\/\/futurenews24.com\/index.php\/2026\/07\/01\/july-2026-newsletter\/"},"author":{"name":"Future News 24","@id":"https:\/\/futurenews24.com\/#\/schema\/person\/cecad1bde21cfc357cf70128144d6c83"},"headline":"July 2026 Publication – by Piers Stobbs","datePublished":"2026-07-01T05:01:00+00:00","dateModified":"2026-07-06T06:59:09+00:00","mainEntityOfPage":{"@id":"https:\/\/futurenews24.com\/index.php\/2026\/07\/01\/july-2026-newsletter\/"},"wordCount":4780,"commentCount":0,"publisher":{"@id":"https:\/\/futurenews24.com\/#organization"},"image":{"@id":"https:\/\/futurenews24.com\/index.php\/2026\/07\/01\/july-2026-newsletter\/#primaryimage"},"thumbnailUrl":"https:\/\/substackcdn.com\/image\/fetch\/$s_!6uEr!,w_1200,h_675,c_fill,f_jpg,q_auto:good,fl_progressive:steep,g_auto\/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F67ad08f8-c41a-4fe4-93fb-072143bc28d7_1604x794.png","keywords":["July","Newsletter","Piers","Stobbs"],"articleSection":["Ethics & Policy"],"inLanguage":"en-US","potentialAction":[{"@type":"CommentAction","name":"Comment","target":["https:\/\/futurenews24.com\/index.php\/2026\/07\/01\/july-2026-newsletter\/#respond"]}]},{"@type":"WebPage","@id":"https:\/\/futurenews24.com\/index.php\/2026\/07\/01\/july-2026-newsletter\/","url":"https:\/\/futurenews24.com\/index.php\/2026\/07\/01\/july-2026-newsletter\/","name":"July 2026 Publication - by Piers Stobbs - Future News 24","isPartOf":{"@id":"https:\/\/futurenews24.com\/#website"},"primaryImageOfPage":{"@id":"https:\/\/futurenews24.com\/index.php\/2026\/07\/01\/july-2026-newsletter\/#primaryimage"},"image":{"@id":"https:\/\/futurenews24.com\/index.php\/2026\/07\/01\/july-2026-newsletter\/#primaryimage"},"thumbnailUrl":"https:\/\/substackcdn.com\/image\/fetch\/$s_!6uEr!,w_1200,h_675,c_fill,f_jpg,q_auto:good,fl_progressive:steep,g_auto\/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F67ad08f8-c41a-4fe4-93fb-072143bc28d7_1604x794.png","datePublished":"2026-07-01T05:01:00+00:00","dateModified":"2026-07-06T06:59:09+00:00","breadcrumb":{"@id":"https:\/\/futurenews24.com\/index.php\/2026\/07\/01\/july-2026-newsletter\/#breadcrumb"},"inLanguage":"en-US","potentialAction":[{"@type":"ReadAction","target":["https:\/\/futurenews24.com\/index.php\/2026\/07\/01\/july-2026-newsletter\/"]}]},{"@type":"ImageObject","inLanguage":"en-US","@id":"https:\/\/futurenews24.com\/index.php\/2026\/07\/01\/july-2026-newsletter\/#primaryimage","url":"https:\/\/substackcdn.com\/image\/fetch\/$s_!6uEr!,w_1200,h_675,c_fill,f_jpg,q_auto:good,fl_progressive:steep,g_auto\/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F67ad08f8-c41a-4fe4-93fb-072143bc28d7_1604x794.png","contentUrl":"https:\/\/substackcdn.com\/image\/fetch\/$s_!6uEr!,w_1200,h_675,c_fill,f_jpg,q_auto:good,fl_progressive:steep,g_auto\/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F67ad08f8-c41a-4fe4-93fb-072143bc28d7_1604x794.png"},{"@type":"BreadcrumbList","@id":"https:\/\/futurenews24.com\/index.php\/2026\/07\/01\/july-2026-newsletter\/#breadcrumb","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https:\/\/futurenews24.com\/"},{"@type":"ListItem","position":2,"name":"July 2026 Publication – by Piers Stobbs"}]},{"@type":"WebSite","@id":"https:\/\/futurenews24.com\/#website","url":"https:\/\/futurenews24.com\/","name":"Future News 24","description":"The Smart Hub for AI and Next-Gen Innovation","publisher":{"@id":"https:\/\/futurenews24.com\/#organization"},"potentialAction":[{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https:\/\/futurenews24.com\/?s={search_term_string}"},"query-input":{"@type":"PropertyValueSpecification","valueRequired":true,"valueName":"search_term_string"}}],"inLanguage":"en-US"},{"@type":"Organization","@id":"https:\/\/futurenews24.com\/#organization","name":"Future News 24","url":"https:\/\/futurenews24.com\/","logo":{"@type":"ImageObject","inLanguage":"en-US","@id":"https:\/\/futurenews24.com\/#\/schema\/logo\/image\/","url":"https:\/\/futurenews24.com\/wp-content\/uploads\/2026\/06\/fn24-favicon.png","contentUrl":"https:\/\/futurenews24.com\/wp-content\/uploads\/2026\/06\/fn24-favicon.png","width":250,"height":250,"caption":"Future News 24"},"image":{"@id":"https:\/\/futurenews24.com\/#\/schema\/logo\/image\/"}},{"@type":"Person","@id":"https:\/\/futurenews24.com\/#\/schema\/person\/cecad1bde21cfc357cf70128144d6c83","name":"Future News 24","image":{"@type":"ImageObject","inLanguage":"en-US","@id":"https:\/\/secure.gravatar.com\/avatar\/d57f07142d73cb5503ab2446ea7bc9ef3d0a5ba378d64a6157692311e42bf097?s=96&d=mm&r=g","url":"https:\/\/secure.gravatar.com\/avatar\/d57f07142d73cb5503ab2446ea7bc9ef3d0a5ba378d64a6157692311e42bf097?s=96&d=mm&r=g","contentUrl":"https:\/\/secure.gravatar.com\/avatar\/d57f07142d73cb5503ab2446ea7bc9ef3d0a5ba378d64a6157692311e42bf097?s=96&d=mm&r=g","caption":"Future News 24"},"sameAs":["https:\/\/futurenews24.com"],"url":"https:\/\/futurenews24.com\/index.php\/author\/mridulpahuja20\/"}]}},"_links":{"self":[{"href":"https:\/\/futurenews24.com\/index.php\/wp-json\/wp\/v2\/posts\/1940","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/futurenews24.com\/index.php\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/futurenews24.com\/index.php\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/futurenews24.com\/index.php\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/futurenews24.com\/index.php\/wp-json\/wp\/v2\/comments?post=1940"}],"version-history":[{"count":1,"href":"https:\/\/futurenews24.com\/index.php\/wp-json\/wp\/v2\/posts\/1940\/revisions"}],"predecessor-version":[{"id":1941,"href":"https:\/\/futurenews24.com\/index.php\/wp-json\/wp\/v2\/posts\/1940\/revisions\/1941"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/futurenews24.com\/index.php\/wp-json\/wp\/v2\/media\/1942"}],"wp:attachment":[{"href":"https:\/\/futurenews24.com\/index.php\/wp-json\/wp\/v2\/media?parent=1940"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/futurenews24.com\/index.php\/wp-json\/wp\/v2\/categories?post=1940"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/futurenews24.com\/index.php\/wp-json\/wp\/v2\/tags?post=1940"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}