Introduction

Everyone is holding their breath, and its with good reason. OpenAI’s most anticipated model might just redefine what we expect from AI altogether.

• Sam Altman´s concerns on existential threats to humanity.
• AI’s power to fundamentally shift the way we interact with the world around us.
• AI in the labour market.
• The end of the human creator economy.
• Sam Altman vs Elon Musk x Trump.

and so much more are included in this week’s issue.

Here’s everything we know (and don’t) about the ChatGPT 5 release—from features and pricing to industry impact and what to expect next. Let’s dive in.

What is ChatGPT 5?

Definition and Purpose - What is It Really?

ChatGPT 5, is the next generation, highly anticipated next iteration of OpenAI’s conversational AI model. While no official version has been released (yet), the GPT-5 release is expected to integrate multimodal capabilities, long-term memory, and advanced reasoning—creating a system that feels more like a cognitive collaborator than a chatbot. Even more interesting, is that GPT-5 release will be powered by Strawberry, a new iteration of OpenAI´s large language models, and Orion, OpenAI´s multimodal AI application.

Who are Strawberry and Orion exactly? Contrary to the titular-anime-character- sounding names, Strawberry and Orion are new AI systems launched by OpenAI that are designed not just to provide quick responses, but to reason and think. Lets get familiar with our terms.

Meet Strawberry!

About Strawberry: Strawberry is not one AI “model”, or program, but is actually a complilation of several, known collectively as o1. All-together these models are intended to answer complex questions. What´s more groundbreaking about Strawberry, is it´s ability for judgement and decision-making, that take it a step closer to human-like intelligence.

The level of reasoning that Strawberry is capable of is still TBD (good example: can it design weapons of mass destruction or bypass chess rules to beat you in a game? We don´t yet know the extent of Strawberry´s risk). Thankfully, OpenAI cites Strawberry models are only “medium risk” for its potential to assist people like scientists in developing chemical, biological, radiological and nuclear weapons. Yippie.

The Powers of Persuasion.

OpenAI evaluated Strawberry as having slightly elevated risk of persuading humans to change their beliefs. Yep, that means that this potential model is more persuasive, and more manipulative than your current ChatGPT therapist.

Meet Orion!

About Orion: Orion has been working with OpenAI since February 27th, 2025 (referred to as GPT 4.5). He´s a Pisces who enjoys a good morning Americano, loves long walks by the beach, and is looking for love…..ok that last part may have been a little extra, BUT Orion is here and ready to mingle.

Orion is believed to be 100 times more powerful than GPT-4, its predecessor. The system is also believed to be a major step toward artificial intelligence that can perform a wide variety of tasks like a human. Orion is essentially built to take the innovations of Strawberry to a broader, more reliable AI infrastructure - reducing hallucinations, improving reliability, and enhancing real-world interaction and autonomy.

The GOATed Fantastic Duo!

So Are They Working Together?

So yes…but no. Strawberry and Orion are not working independently, nor are they competing against each other. Rather, the plan for GPT-5, is for Strawberry to operate as a core reasoning module within Orion/GPT-5. While Strawberry can handle highly technical research questions, Orion uses this output as part of a broader conversation or project- learning and improving from each interaction. Strawberry manages nuanced inquiries with high precision, while Orion oversees, refines, escalates and integrates those results into workflows, scientific research or customer service. The perfect synergy.

Evolution From Previous Versions - Lets Look At the Timeline

Ok…So What Does All That Actually Mean??

Let's break it down in human-speak. The changes awaiting users in the coming GPT-5 release, simply cannot be underscored.

For the vast majority of users, the model selector tab is not something that they are even aware of. From the infancy days of ChatGPT-1, all the way up to the current 4o model, the average person will know ChatGPT as well...just ChatGPT without any added frills.

And why bother going through each individual model? Well, the truth is that each iteration of ChatGPT has a unique set of capabilities that sets it apart from the others. Hey, even twins have their own tells for distinction. The fact is — not all ChatGPT's are built alike. So what about GPT-5 release is so hype-worthy?

Think of this release as the ball-drop we experienced going from the yearn 1999 to 2000. We are entering a new millennium of tech capability. Not only is this new model the culmination of the best slices of past models, its essentially an entire new model that will be unifying existing systems of LLM.

This is incredibly exciting news! Shall we jump into the champagne popping? Well, not everyone is feeling so enthusiastic about the release.

For Context

Sam Altman (CEO of OpenAI), isn't quite in the mood for celebrating the ChatGPT-5 release. If anything, he is sounding the alarm.

During Altman's podcast appearance on This Past Weekend with Theo Von, Altman attested that the ChatGPT-5 release could have massive implications for the way that we interact with our world. He spoke about the dangers of relying on AI agents when it comes to mental health, and the uncharted waters of letting children get their easy-access dopamine hits through agent scrolling and social media interaction. Altman said:

There are moments in the history of science, where you have a group of scientists look at their creation and just say, you know: what have we done?

This concern is voiced at a time when Geoff Lewis, one of the managing partners at Bedrock (a multi-billion dollar investment firm, which backs some insanely high-profile tech companies like OpenAI and Vercel), posted a disturbing video on X that has since, shocked the world. Either a call for the complete restructure of AI systems that are so-casually used by everyone and their grandmother (literally) or a cry for the media to shine a spotlight on emerging mental health crises as a result of AI agents. Only time will tell if we are bearing witness to the solution to our epoch's greatest problems - or a new kind of Frankenstein's monster that cannot be contained.

Some Further Food for Thought in the Tech Mayhem: The Trump Story

Back in January 2025, President Donald Trump announced the formation of a colossal joint venture anointed “Stargate” with OpenAI, Oracle, and SoftBank, pledging a lofty $100 billion upfront and scaling up to $500 billion over four years. The focus remained on building massive AI infrastructures—first and foremost data centers that were essential to powering large language models and AI applications.

At the same time, Trump actively supported Elon Musks's Starlink satellite internet service. There are still some substantial debates about whether moves like policy changes in broadband infrastructure (ahem-like overhauling $42 billion federal grand program to be “tech neutral”) to benefit Starlink, were due to Musks's connection to Trump as an ally during his everything-but-smooth political campaign.

The takeaway? While Trump supports Musk and Starlink, Musk himself is no longer a participant at the Stargate venture. In fact, Musk publicly questioned (gasp!) whether the consortium (OpenAI, Oracle, SoftBank) truly had access to the full $500 billion promised by Trump, suggesting a deeper dive into skepticism, and the further widening ravine between Altman and OpenAI.

By now, you have the full context.

But lets get into the meat of what these changes will mean for you.

Expected Features of ChatGPT 5 Release

New Capabilities and Improvements—What You Can Expect!

💡Persistent Memory = No More Repeating Yourself
Past Models were kind of like goldfish when it came to memory that mattered. GPT-5 remembers your preferences, past convos, tone of voice, and even what kind of content is your cup of tea.

🧠 Unified Reasoning = Getting Smarter, Context-Aware Replies
Instead of surface-level responses, GPT-5 connects the dots across tasks, formats, and goals like a true collaborator.

🎥 Full Multimodal Input = Going Beyond Text and Images
This model will handle text, voice, video, and visuals seamlessly. Expect voice commands and responses, video understanding and summarization and dynamic interactions (e.g., describe this screenshot or summarize this clip).

🎤 Agent Mode = An Autopilot for Your To-Do's
The agent feature is already live with the ChatGPT-4 version, so what can we expect? Well, for the agent to perform and automate tasks for you using its own computer, handling tasks from start to finish (browse websites, run code, fill out forms, and actually complete multi-step tasks autonomously).

Pricing Model for ChatGPT 5´s Release

While a formal price for the GPT-5 release hasn´t been announced by OpenAI, ChatGPT Strawberry (which OpenAI refers to as ChatGPT 01) is already available in beta form at no additional charge to those who already have a $20 USD monthly subscription.

So a limited sneak-peek at one component of GPT-5 —Strawberry— is cheap or borderline free-ish.

For the whole ChatGPT 5.o package, theres a hefty change in price. What´s known is that GPT-5 can cost businesses as much as $2,000 USD monthly. If this holds true for rollout, this new model will be beyond the reach of most individual users.

As we wait for confirmation we can confirm to expect the standard pricing structure, as OpenAI announced GPT-5 will be available in multiple pricing tiers:

• Free Tier – Unlimited chat access at a standard intelligence level, with possible limits to prevent abuse.
• Plus Tier – Higher intelligence and better performance, similar to how GPT-4o was offered in ChatGPT Plus.
• Pro Tier – Full access to GPT-5 at its highest intelligence level, along with OpenAI’s most advanced tools and integrations.

While speculation for pricing he question continues to float, we´re curious about just how many, and whether any users will be willing to go above the current ChatGPT Pro fee of $200 USD.

Size Matters: Just How Big is GPT-5?

Size matters where it counts. Let´s talk about tokens.

Ok enough with the jokes, lets get to understand why tokens are such a big deal when it comes to the scale of the ChatGPT-5 release. Ilya Sutskever, one of the co-founders and former Chief Scientist at OpenAI, emphasized the role of tokens.

The scale of data (i.e, the number of tokens ithe model is trained on) is oneof the primary drivers of performance in large language models.—Sutskever

Simply put:

• Training on more tokens leads to better generalization (up to a point that is).

In the world of AI, the size of a model— measured in parameters and training tokens- directly impacts its intelligence and capabilities. Essentially, parameters are the internal settings (like weights) that the model learns from data to make predictions, while tokens are individual units of text (words, characters, etc.) that the model processes during training.

These components define the brain power of the systems.

Larger models like Claude 3 Opus (with an estimated IQ of 160 based on GPQA scores, and GPT-4o can solve graduate- level problems far beyond average human reasoning. These GPQA benchmarks (Graduate-level Problem and Question Answering) reveal just how “smart” AI has become: as you can see in the chart, models scoring 50% are operating in the top 0.2% of human intelligence.

This scale is so important to understand the giant leap expected of the ChatGPT-5 release. For reference, this is what we´re looking at when we compare between GPT-4 and GPT-5.

So just for reference, this graphic compares the GPU scale (compute budget) used to train:

• GPT-4 👉75 billion units
• GPT-5 👉10 trillion units

Thats a little jump of 133x increase in computing!

Why This Matters for Performance

Simply put: more GPU´s = more computer power = bigger brains = more capability. Here´s why thats a key win in this story.

1. More Data + More Parameters = More Intelligence

• GPT-4 was already brilliant (GPQA= 53.6%, IQ = 146)
• BUT, GPT-5 already trained 10 trillion GPU operations, is expected to handle longer context, deeper reasoning, and multi-step planning at superhuman levels.

1. Training Budget Dictates Model Size

• The bigger the better holds true, the more budget and training you compute to an AI system, the more complex and more nuanced it can become.

1. Breakthroughs in Reasoning, Memory and Autonomy

• This jump could bring persistent memory (thats long-term knowledge of you!), unified reasoning across modalities, and Agent Mode- allowing GPT-5 to act more like an assistant that complete complex tasks autonomously.

Is ChatGPT-5 going to be better than Grok?

Lets address the giant elephant in the room, aka the questions we´re all asking: who´s winning the AI race between Musk and Altman? Will it Grok or ChatGPT-5 that will come out on top?

According to the Latest Grok Benchmarks vs Competitors: Grok 4 Dominates AI Performance Rankings

Musk´s own xAi launched Grok 4 in July 2025. The baby AI system is pulverizing the competition, with benchmark results that have “PhD-level” performance across academic domains and has set multiple new records on challenging benchmarks.

Breakthrough Performance on Humanity´s Last Exam

Grok 4 is outpacing humans without questions. Nowhere is it more evident than in its results from Humanity´s Last Exam (HLE), a skull- crushing 2,500 question benchmark designed to test AI at the frontier of human knowledge agress STEM and humanities. The results came in:

• Grok 4 Heavy with tools = 44.4%
• Gemini 2.5 Pro with tools = 26.9%
• OpenAI o3 with tools: 24.9%

Compare and Contrast with the Competition

The numbers are in, and Grok 4´s heavy scoring is nearly double its closest competitor. What does this mean for its capability?

• Advanced Maths and Formal Logic: Grok 4 matches graduate- level problem sets and sits just behind OpenAI´s o-series competition.
• Coding: The model still lags Opus on pass-at 1 Python but beats Gemini and leaves open-weight models in the dust.
• Context and Memory: Geminin´s and GPT-4.1´s 1M tokens remain unmatched, however Grok´s 256K remains decent in the game. Real world test show stable retrieval across > 200K-token transcripts.
• Speed and Price: Roughly 75 output tokens/s - noticeable slower than o3 but still faster than Gemini.

Grok- Controversy in Media

Making splashes as always, Musks latest endeavour is already splitting hairs amongst critics in the media.

CNN Business reports that in the days just before Grok 4´s launch, Grok 3 produced inflammatory and conspiratorial antisemitic rants on X following a system prompt update encouraging “politically incorrect” but “well substantiated” statements. Yikes.

What else will we see in the comparison race in the days ahead? Only time will tell. One thing is for certain. Musk´s flair for the dramatic and controversial may be the cherry- on-top for Altman´s win, not only the triumph of GPU volume promised by GPT-5, but in overall public perception of the release.

Regional Availability

Again, no official confirmation regions yet, BUT we can expect the ChatGPT-5 release rollout to follow the following pattern:

• U.S., UK, Canada likely first
• EU rollout dependent on AI Act compliance

ChatGPT-5 Release Date

Official Release Date- TBD

Everyone hold on to your horses. As of now, there is no confirmed release date. The release date is so hush-hush, that in July 2025 in a Reddit AMA, Sam Altman explicitly stated “Nothing that we are going to call GPT-5” when asked about releases this year. And yet, here we are in August awaiting the latest drop. Expectant users can anticipate a full rollout for the drop when it does happen. We can all look forward to a mini and nano versions tailored for different workloads, with ChatGPT-5 release being projected to land on both ChatGPT and the API, while the leaner nano version looks to stay API- only.

So officially, OpenAI has not published a formal timeline for the GPT-5 release, citing concerns over safety, readiness, and computational viability. Leaving the question: why all this hype and none of the deliverables? Lets look into it.

The Tea on Speculations and Rumors

Initial speculation suggested a mid-2024 release, but reports of two failed $500M training runs have cast doubt on near-term deployment. Regulatory shifts—like the EU AI Act—may also account for slowed regional rollouts.

But Lets Back Up, and Dive Into the Juicy Stuff: The OpenAI Feud for World Domination and the Feud to End All Feuds.

For those that don´t know- the shakedown between Sam Altman and the board members of OpenAI, was one for the history books. We´ll ge to Elon Musk in a minute. The stakes for the shakedown? Nothing as mediocre as power-struggles for corporate ownership- but for the future of humanity itself. Back in November 2023, the world was covertly living in an apocalyptic version of Succession. After weeks of secret calls and under the table deals, the OpenAI Board reached a consensus. Altman got the boot.

The Drama didn´t stop flowing there. After his layoff, Altman´s most loyal staff resigned, while others signed an open letter calling for his return to office. An army of investors, including the heaviest artillery, Microsoft, were starting to get scared. Without the leading talent of Altman at the helm of OpenAI, there were doubts that the company could continue to survive. Needless to say, some board members who were directly involved in the decision to fire Altman, promptly switched sides in a matter of days. Altman was back to holding his crown once again as CEO in a matter of days.

But the crown did not rest easy on his head from the get go. In fact- lets trace back to the heritage roots of OpenAI and put some key names to light. The company was originally co- founded by Elon Musk, along with Sam Altman, Ilya Sutskever, Greg Brockman, and others in 2015, as a non-profit AI research company. The funds backing the now-powerhouse, was initially provided by Musk.

This is where things get interesting….

While OpenAI initially branched off as a nonprofit company, it later launched a for-profit subsidiary to oversee its commercial operations. Its product (the humble ChatGPT) already had a million users within days of its release in 2022, and 100 million within a couple months. Quickly becoming the fastest growing consumer app in history. But with stiff competition heating up from giants like Google and Amazon, OpenAI needed to raise more money- and do it quickly. To make the process easier, it ventured off on an a new course: a for-profit subsidiary to a public benefit corporation. Problem solved.

Well, not so much. The move rubbed key investors like Musk the wrong way. Musk himself invested more than $44 million between 2016 to 2020 in OpenAI, under the belief it would remain a nonprofit. Sticking to its initial promise of serving the greater good of humanity, and not become shackled to shareholders concerned with profit margins.

So what did Musk allege as the two Lords of Technology duked it out in court?

Musk stated the following in a court record filed last year:

OpenAI´s conduct could have seismic implications for Silicon Valley and, if allowed to stand, could represent a paradigm shift for technology start-ups.

In court, Musk alleged that OpenAI veered ay from its founding principles, citing his significant investment, help in recruiting top talent under that premise and his servitude as co- chair (not to mention him coming up with the name “OpenAI”).

So what did OpenAI respond with? According to the company, Musk´s lawsuit is an attempt to slowdown a rival since he left the board in February 2018 and became a direct competitor by launching xAI which developed the AI assistant Grok (a little petty but hey, thats business).

Since then, Musk has been actively trolling OpenAI (cleverly dubbed “Scam Altman” by himself) on social media for keeping its source code private, and signing a widely publicized manifesto calling for a pause in AI development, even as he pursued his own AI projects. How this man has time for sleep- don´t ask.

So What Does All This Drama Mean for the Release of ChatGPT-5?

While there´s no direct impact on release timing for GPT-5, the level of petty competition and antagonism, have arguable moved OpenAI to push forward more aggressively, rather than the slow down on release which Musk may have been banking on. The delicious bouts of online snubbery have fueled headlines, drawing attention to the stakes of advanced AI. Amidst this battle of titans, the urgency of the Altman vs Musk face-off has raised the release of GPT-5 to the global arena. Both tech and society, are eagerly watching.

Comparison with Previous Release Dates- What to Expect for You

OpenAI’s Announcement- What You Need to Hear

Key Messages from Leadership

Altman and OpenAI leadership have jointly emphasized that when unleashing GPT-5 on the world the priority is given to:

• Safety and alignment over speed
• A pause on massive scale-ups in favor of economic and architectural innovation

Having said that, Altman is himself not without his doubts. Referring back to the infamous Theo Von´s This Past Weekend podcast, Altman confessed that testing GPT-5 left him with a sense of unease. When witnessing the model solve a complex problem that he couldn´t solve, Altman stated:

I felt useless…

And that´s coming from one of the creators of the product. Altman´s comparison of the new release to the Manhattan Project- in case you haven´t watched Oppenheimer: it was the top-secret U.S. effort during WWII that led to the creation of the atomic bomb. This reveals just how close we are at the junction of tech merging into a Pandora´s box that could have tremendous consequences on society.

This leads into ethical considerations for AI.

Altman´s statement reflects the broader concern amongst AI leaders: is progress outpacing precaution? Not only did his creation accelerate human productivity and automate work, it now has the potential to disrupt social structures, destroy trust in information (hello Dead Internet Theory), and ask the question: what´s so unique about humans if machines are already taking over?

Are We Looking at the Death of the Human Creator Economy?

FYI, the global creator economy market is worth over 125 billion USD. That´s a lot. What is the implication of this content becoming entirely produced by AI systems?

Creator Dan Koe, voiced in February 2025, that over consumption of online content leads to psychic entropy of the mind. In simple terms, the more brainrot you consume online leads you to…..well, rot. With the average American spending around 7 hours of the day online, there´s reason to be concerned for when automated content creation powered by AI induces chaos.

To that extent Altman says:

In a future (that is AI run), with AI leading tech developments and medical innovations, I think that what people want at the end of the day- is to have the opportunity to co-create their future with AI.

While its true that Pandora´s box is waiting to be opened, there is no denying that these changes can bring about the next golden age… for those who know how to leverage its opportunities and stay informed. 👉To stay up to date on all things Tech, AI and Upskilling your life today check out Coursiv

Guidelines for Optimal Usage: Your How-To Guide

So far, we´ve covered a lot of ground. So let´s move into best practices. How can you best use GPT-5 for when the ball drops? Lets get into it.

What Not to Tell ChatGPT

Always treat AI as a public-facing tool. Use incognito mode for sensitive topics (yes, prompting how to get away with tax evasion never ends well), and never input confidential data without knowing the platform’s terms.

Five Things to Avoid Sharing

1. Personal identification (e.g., SSN, credit cards)
2. Confidential business data
3. Medical records
4. Legal case details
5. Children’s information

Additional food for thought: Best Practice Implementation

• Start with Clear Objectives: define specific use cases before implementation & establish measurable success metrics.
• Ensure Data Privacy and Security: Check your safety settings regularly for compliance and safety checks.
• Build AI Literacy: Train yourself and/or your team on the latest AI capabilities and best practices to stay ahead of the curve.

Frequently Asked Questions About the ChatGPT 5 Release

Q) How much better will ChatGPT 5 be?

Smarter, more contextual, memory-based, and multimodal—but slower and likely more expensive.

Q) Will GPT-5 be free?

Unlikely. Expect Pro-only or usage-gated access.

Q) What is the IQ of GPT-5?

Estimates suggest IQ ~155. Take AI IQ benchmarks with a grain of salt.

Q) Will GPT-5 have more memory?

Yes. It will offer persistent memory with user controls.

Q) What will ChatGPT 5 be able to do?

Multimodal reasoning, autonomous task handling, real-time collaboration, and advanced personalization.

Q) How much bigger is GPT-5 than 4?

Likely 2–3x larger in parameters and context capacity.

Q) What will ChatGPT 6 be able to do?

Real-time agents, cloud-based persistent memory, wearables integration, and full autonomy.

Q) How many parameters are in GPT-6?

Estimates range from 1–2 trillion+ depending on infrastructure.

✅ Let´s Wrap it Up

The GPT-5 release represents more than just an upgrade—Fit’s a paradigm shift in how we’ll work, learn, and build in an AI-powered future. But to take full advantage of its power, users need more than access—they need AI fluency.

At Coursiv, we’re helping professionals and beginners build those skills today—with Tailored Courses, and Daily Challenges on how to prompt, collaborate, and thrive in the age of reasoning AI.

👉 Explore Coursiv Today