Investors have been pouring billions into AI companies over the past few years, as the technology continues to hold sway in the Valley and thus the world. But not all AI companies are grabbing investor attention.
Indeed, even as it seems every company these days is rebranding to include “AI” in its name, some startup ideas are just no longer in favor with investors. TechCrunch spoke with VCs to learn what investors aren’t looking for in AI software-as-a-service startups anymore.
Popular SaaS categories for investors now include startups building AI-native infrastructure, vertical SaaS with proprietary data, systems of action (those helping users complete tasks), and platforms deeply embedded in mission-critical workflows, according to Aaron Holiday, a managing partner at 645 Ventures.
But he also gave a list of companies that are considered quite boring to investors these days: Startups building thin workflow layers, generic horizontal tools, light product management, and surface-level analytics — basically, anything an AI agent can now do.
Abdul Abdirahman, an investor at the firm F Prime, added that generic vertical software “without proprietary data moats” is no longer popular, and Igor Ryabenky, a founder and managing partner at AltaIR Capital, went deeper on that point. He said investors aren’t interested in anything, really, that doesn’t have much product depth.
“If your differentiation lives mostly in UI [user interface] and automation, that’s no longer enough,” he said. “The barrier to entry has dropped, which makes building a real moat much harder.”
New companies entering the market now need to build around “real workflow ownership and a clear understanding of the problem from day one,” he said. “Massive codebases are no longer an advantage. What matters more is speed, focus, and the ability to adapt quickly. Pricing also needs to be flexible: rigid per-seat models will be harder to defend, while consumption-based models make more sense in this environment.”
Techcrunch event
San Francisco, CA | October 13-15, 2026
Jake Saper, a general partner at Emergence Capital, also had thoughts on ownership. To him, the differences between Cursor and Claude Code are the “canary in the coal mine.”
“One owns the developer’s workflow, the other just executes the task,” Saper continued. “Developers are increasingly choosing the execution over process.”
He said any product dealing with “workflow stickiness” — meaning trying to attract as many human customers as possible to continuously use the product — might find themselves in an uphill battle as agents takeover the workflow.
“Pre-Claude, getting humans to do their jobs inside your software was a powerful moat, but if agents are doing the work, who cares about human workflow?” he told TechCrunch.
He also thinks integrations are becoming less popular, especially as Anthropic’s model context protocol (MCP) makes it easier than ever to connect AI models to external data and systems. This means someone doesn’t need to download multiple integrations or build their own customer integrations; they can just use the MCP.
“Being the connector used to be a moat,” Saper said. “Soon, it’ll be a utility.”
Also, no longer en vogue include the “workflow automation and task management tools that enable the coordination of human work become less necessary if, over time, agents just execute the tasks,” Abdirahman said, citing examples, mainly public SaaS companies whose stocks are down as new AI-native startups arise with better, more efficient technology.
Ryabenky said the SaaS companies struggling to raise right now are the ones that can easily be replicated, he said.
“Generic productivity tools, project management software, basic CRM clones, and thin AI wrappers built on top of existing APIs fall into this category,” he said. “If the product is mostly an interface layer without deep integration, proprietary data, or embedded process knowledge, strong AI-native teams can rebuild it quickly. That is what makes investors cautious.”
Overa, what remains attractive about SaaS is depth and expertise, with tools embedded in critical workflows. He said companies should right now look into integrating AI deeply into their products and update their marketing to reflect that, Ryabenky continued.
“Investors are reallocating capital toward businesses that own workflows, data, and domain expertise,” Ryabenky said. “And away from products that can be copied without much effort.”
I did not expect Running With Scissors to announce a new FPS where you play as a victim of one of the massacres depicted in the original Postal. Then again, I find it wise not to expect anything from Running with Scissors. The developer made its name on stoking outrage with its largely terrible games, though I will concede that Postal 2 was, despite its many, many flaws, ahead of its time in certain ways.
Even so, Flesh & Wire is a twist I didn’t see coming. Described by Running With Scissors as a “horror first-person shooter”, it puts you in the shoes of Angel, a young woman seeking vengeance against the Postal Dude after being caught on the wrong end of his rifle during his 1997 killing spree. The game will involve tracking the Dude across the Arizona desert, battling military forces and your own traumatic memories as you search for vengeance.
Running With Scissors released a trailer for the game to go with the announcement, and on the face of it, Flesh & Wire certainly seems to be striving for a more serious tone than the mainline Postal series, though the combat looks like it needs tightening up even in what little we see of it. I’m also not convinced it’s going to be the most nuanced depiction of struggles with PTSD, given how Angel wraps her arms in barbed wire and can apparently employ supernatural forces to fight enemies.
Flesh & Wire – Official Announcement Trailer | IGN Fan Fest 2026 – YouTube
There’s no doubt that wrangling with the ramifications of the original Postal’s story is an interesting idea, especially given the US’ torrid history with mass shootings in the years since. Do I trust Running with Scissors in handling such material? No! And neither should you!
Nonetheless, I would love to be proven wrong and for Flesh & Wire to turn out great. At the very least, Running With Scissors has time to get the game right. Flesh & Wire isn’t coming out until sometime next year. If in the meantime you want to play a Postal game that doesn’t suck, Hyperstrange’s spinoff Postal: Brain Damaged is an entertaining, if thoroughly disposable, boomer shooter lark.
Keep up to date with the most important stories and the best deals, as picked by the PC Gamer team.
Modern generative‑AI experiences hinge on speed. When a user types a question into a chatbot or triggers a long‑form summarization pipeline, two latency metrics define their experience: Time‑to‑first‑token (TTFT) and throughput. TTFT measures how quickly the first sign of life appears after a prompt; throughput measures how many tokens per second, requests per second or other units of work a system can process. Over the past two years, these metrics have become central to debates about model selection, infrastructure choices and user satisfaction.
In early generative systems circa 2021, any response within a few seconds felt magical. Today, with LLMs embedded in IDEs, voice assistants and decision support tools, users expect nearly instantaneous feedback. New research on goodput—the rate of outputs that meet latency service‑level objectives (SLOs)—shows that raw throughput often hides poor user experience. At the same time, innovations like prefill‑decode disaggregation have transformed server architectures. In this article we unpack what TTFT and throughput actually measure, why they matter, how to optimize them, and when one should take priority over the other. We also weave in Clarifai’s platform features—compute orchestration, model inference, local runners and analytics—to show how modern tooling can support these goals.
Quick Digest
Definitions & Evolution: TTFT reflects responsiveness and psychological perception, while throughput reflects system capacity. Goodput bridges them by counting only SLO‑compliant outputs.
Context‑Driven Trade‑offs: For human‑centric interfaces, low TTFT builds trust; for batch or cost‑sensitive pipelines, high throughput (and goodput) drives efficiency.
Optimization Frameworks: The Perception–Capacity Matrix, Acknowledge‑Flow‑Complete model and Latency–Throughput Tuning Checklist provide structured approaches to balancing metrics across workloads.
Clarifai Integration: Clarifai’s compute orchestration and local runners reduce network latency and support hybrid deployments, while its analytics dashboards expose real‑time TTFT, percentile latencies and goodput.
Defining TTFT and Throughput in LLM Inference
Why do these metrics exist?
The labels may be new, but the tension behind them is old: systems must feel responsive while maximizing work done. TTFT is defined as the time between sending a prompt and receiving the first output token. It captures user‑perceived responsiveness: the moment a chat UI streams the first word, anxiety diminishes. Throughput, in contrast, measures total productive work—often expressed as tokens per second (TPS) or requests per second (RPS). Historically, early inference servers optimized throughput by batching requests and filling GPU pipelines; however, this often delayed the first token and undermined interactivity.
How are they calculated?
At a high level, end‑to‑end latency equals TTFT + generation time. Generation time itself can be decomposed into time‑per‑output‑token (TPOT) and the total number of output tokens. Throughput metrics vary: some frameworks compute request‑weighted TPS, while others use token‑weighted averages. Good instrumentation logs each event—prompt arrival, prefill completion, token emission—and counts tokens to derive TTFT, TPOT and TPS.
Metric
What it measures
Core formula
TTFT
Delay until first token
Arrival → First token
TPOT / ITL
Average delay between tokens
Generation time ÷ tokens generated
Throughput (TPS)
Tokens processed per second
Tokens ÷ total time
Goodput
SLO‑compliant outputs per second
Sum of outputs meeting SLO / total time
Trade‑offs and misinterpretations
Low TTFT delights users but can limit throughput because smaller batches underutilize GPUs. Conversely, maximizing throughput via large batches or heavy prompts can inflate TTFT and degrade perception. A common mistake is to equate average latency with TTFT; averages hide long‑tail percentiles that frustrate users. Another misconception is that high TPS implies good user experience; in reality, a provider may produce many tokens quickly but start streaming after several seconds.
Original Framework: Perception–Capacity Matrix
To help teams visualize these dynamics, consider the Perception–Capacity Matrix:
Quadrant I:High TTFT / Low Throughput – worst of both worlds; often due to large prompts or overloaded hardware.
Quadrant II:Low TTFT / Low Throughput – ideal for chatbots and code editors; invests in quick response but processes fewer requests concurrently.
Quadrant III:High TTFT / High Throughput – batch‑oriented pipelines; acceptable for long‑form generation or offline tasks but poor for interactivity.
Quadrant IV:Low TTFT / High Throughput – aspirational; often requires advanced caching, dynamic batching and disaggregation.
Mapping workloads onto this matrix helps decide where to invest engineering effort: interactive applications should target Quadrant II, while offline summarization can live in Quadrant III.
Expert Insights
Interactive applications depend on TTFT: Anyscale notes that interactive workloads benefit most from low TTFT.
Throughput shapes cost: Larger batches and high TPS maximize GPU utilization and lower per‑token cost.
High TPS can be misleading: Independent benchmarks show providers with high TPS but poor TTFT.
Clarifai analytics: Clarifai’s dashboard tracks TTFT, TPOT and TPS in real time, enabling users to monitor long‑tail percentiles.
Quick Summary
What is TTFT? The time until the first token appears.
Why care? It shapes user perception and trust.
What is throughput? Total work done per second.
Key trade‑off: Low TTFT usually reduces throughput and vice versa.
Why TTFT Matters More for Human‑Centric Applications
Humans hate waiting in silence
Psychologists have shown that people perceive idle waiting as longer than the actual time. In digital interfaces, a delay before the first token triggers doubts about whether a request was received or if the system is “stuck.” TTFT functions like a typing indicator—it reassures the user that progress is happening and sets expectations for the rest of the response. For chatbots, voice assistants and code editors, even 300 ms differences can affect satisfaction.
Operational playbook to reduce TTFT
Measure baseline: Use observability tools to collect TTFT, p95/p99 latencies and GPU utilization; Clarifai’s dashboard provides these metrics.
Optimize prompts: Remove unnecessary context, compress instructions and order information by importance.
Choose the right model: Smaller models or Mixture‑of‑Experts configurations shorten prefill time; Clarifai offers small models and custom model uploads.
Reuse KV caches: When repeating context across requests, reuse cached attention values to skip prefill.
Deploy closer to users: Use Clarifai’s Local Runners to run inference on‑premise or at the edge, cutting network delays.
For chatbots and real‑time translation, aim for TTFT under 500 ms; code completion tools may require sub‑200 ms latencies.
When TTFT should not be prioritized
Batch analytics: If responses are consumed by machines rather than humans, a few seconds of TTFT have minimal impact.
Streaming with heavy generation: In tasks like essay writing, users may accept a slower start if tokens subsequently stream quickly. However, avoid using long prompts that block user feedback for tens of seconds.
Network noise: Optimizing model-level TTFT doesn’t help if network latency dominates; on‑premise deployment solves this.
Original Framework: Acknowledge‑Flow‑Complete Model
This model breaks user experience into three phases:
Acknowledge – the first token signals the system heard you.
Complete – the answer finishes when the last token arrives or the user stops reading.
By instrumenting each phase, engineers can identify where delays occur and target optimizations accordingly.
Expert Insights
Human reading speed is limited: Baseten notes that humans read only 4–7 tokens per second, so extremely high throughput does not translate to better perception.
TTFT builds trust: CodeAnt highlights how quick acknowledgment reduces cognitive load and user abandonment.
Clarifai’s Reasoning Engine benchmarks: Independent benchmarks show Clarifai achieving TTFT of 0.32 s with 544 tokens/s throughput, demonstrating that good engineering can balance both.
Quick Summary
When to prioritize TTFT? Whenever a human is waiting on the answer, such as in chat, voice or coding.
How to optimize? Measure baseline, shrink prompts, pick smaller models, reuse caches and reduce network hops.
When Throughput Takes Priority—Scaling for Efficiency and Cost
Throughput for batch and server efficiency
Throughput measures how many tokens or requests a system processes per second. For batch summarization, document generation or API backends that process thousands of concurrent requests, maximizing throughput reduces per‑token cost and infrastructure spend. In 2025, open‑source servers began to saturate GPUs by continuous batching, grouping requests across iterations.
Operational strategies
Dynamic batching: Adjust batch size based on request lengths and SLOs; group similar length prompts to reduce padding and memory waste.
Prefill‑decode disaggregation: Separate prompt ingestion (prefill) from token generation (decode) across GPU pools to eliminate interference and enable independent scaling.
Compute orchestration: Use Clarifai’s compute orchestration to spin up compute pools in the cloud or on‑prem and automatically scale them based on load.
Goodput tracking: Measure not just raw TPS but the fraction of requests meeting SLOs.
Decision logic
If tasks are offline or machine‑consumed: Maximize throughput. Choose larger batch sizes and accept TTFT of several seconds.
If tasks require mixed human/machine consumption: Use dynamic strategies; maintain moderate TTFT (<3 s) while increasing throughput via disaggregation.
If tasks are highly interactive: Keep batch sizes small and avoid sacrificing TTFT.
Original Framework: Batch‑Latency Trade‑off Curve
Visualize throughput on one axis and TTFT on the other. As batch size increases, throughput climbs quickly then plateaus, while TTFT increases roughly linearly. The “sweet spot” lies where throughput gains begin to taper yet TTFT remains acceptable. Overlays of cost per million tokens help teams choose the economically optimal batch size.
Common mistakes
Chasing throughput without goodput: Systems that achieve high TPS with many long‑running requests may violate latency SLOs, lowering goodput.
Comparing TPS across providers blindly: Throughput numbers depend on prompt length, model size and hardware; reporting a single TPS figure without context can mislead.
Ignoring data transfer: Throughput gains vanish if network or storage bottlenecks throttle token streaming.
Expert Insights
Research on prefill‑decode disaggregation: DistServe and successor systems show that splitting phases enables independent optimization.
Clarifai’s Local Runners: Running inference on‑prem reduces network overhead and allows enterprises to select hardware tuned for throughput while meeting data residency requirements.
Goodput adoption: Papers published in 2024–2025 argue for focusing on goodput rather than raw throughput, signalling an industry shift.
Quick Summary
When to prioritize throughput? For batch workloads, document pipelines, and scenarios where cost per token matters more than immediate responsiveness.
How to scale? Apply dynamic batching, adopt prefill‑decode disaggregation, track goodput and leverage orchestration tools to adjust resources.
Watch out for: High throughput numbers with low goodput; ignoring latency SLOs; not considering network or storage bottlenecks.
Balancing TTFT and Throughput—Decision Frameworks and Optimization Strategies
Understanding the inherent trade‑off
LLM serving involves balancing two competing goals: keep TTFT low for responsiveness while maximizing throughput for efficiency. The trade‑off arises because prefill operations consume GPU memory and bandwidth; large prompts produce interference with ongoing decodes. Effective optimization therefore requires a holistic approach.
Step‑by‑step tuning guide
Collect baseline metrics: Use Clarifai’s analytics or open‑source tools to measure TTFT, TPS, TPOT and percentile latencies under representative workloads.
Tune prompts: Shorten prompts, compress context and reorder important information.
Select models strategically: Small or Mixture‑of‑Experts models reduce prefill time and can maintain accuracy for many tasks. Clarifai allows uploading custom models or selecting from curated small models.
Leverage caching: Use KV‑cache reuse and prefix caching to bypass expensive prefill steps.
Apply dynamic batching and prefill‑decode disaggregation: Adjust batch sizes based on traffic patterns and separate prefill from decode to improve goodput.
Deploy near users: Choose between cloud, edge or on‑prem deployments; Clarifai’s Local Runners enable on‑prem inference for low TTFT and data sovereignty.
Iterate using metrics: Set SLO thresholds (e.g., TTFT <500 ms, TPOT <50 ms) and iterate. Use Clarifai’s alerting to trigger scaling or adjust batch sizes when p95/p99 latencies exceed targets.
Decision tree for different workloads
Interactive with short responses: Choose small models and small batch sizes; reuse caches; scale horizontally when traffic spikes.
Long‑form generation with human readers: Accept TTFT up to ~3 s; focus on stable inter‑token latency; stream results.
Offline analytics: Use large batches; separate prefill and decode; aim for maximum throughput and high goodput.
Original Framework: Latency–Throughput Tuning Checklist
To operationalize these guidelines, create a checklist grouped by categories:
Prompt Design: Are prompts short and ordered by importance? Have you removed unnecessary examples?
Model Selection: Is the chosen model the smallest model that meets accuracy requirements? Should you switch to a Mixture‑of‑Experts?
Caching: Have you enabled KV‑cache reuse or prefix caching? Are caches being transferred efficiently?
Batching: Is your batch size optimized for current traffic? Do you use dynamic or continuous batching?
Deployment: Are you serving from the region closest to users? Could local runners reduce network latency?
Monitoring: Are you measuring TTFT, TPOT, TPS and goodput? Do you have alerts for p95/p99 latencies?
Reviewing this list before each deployment or scaling event helps maintain performance balance.
Expert Insights
Infrastructure matters: DBASolved emphasizes that GPU memory bandwidth and network latency often dominate TTFT.
Prompt engineering is powerful: CodeAnt provides recipes for compressing prompts and reorganizing context.
Adaptive batching algorithms: Research on length‑aware and SLO‑aware batching reduces padding and out‑of‑memory errors.
Quick Summary
How to balance both metrics? Collect baseline metrics, tune prompts and models, apply caching, adjust batches, choose deployment location and monitor p95/p99 latencies.
Framework to use: The Latency–Throughput Tuning Checklist ensures no optimization area is missed.
Key caution: Over‑tuning for one metric can starve another; use metrics and decision trees to guide adjustments.
Case Study – Comparing Providers & Clarifai’s Reasoning Engine
Benchmarking landscape
Independent benchmarks like Artificial Analysis evaluate providers on common models (e.g., GPT‑OSS‑120B). In 2025–2026, these benchmarks surfaced surprising differences: some providers delivered exceptionally high TPS but had TTFTs above four seconds, while others achieved sub‑second TTFT with moderate throughput. Clarifai’s platform recorded TTFT of ~0.32 s and 544 tokens/s throughput at a competitive cost; another test found 0.27 s TTFT and 313 TPS at $0.16/1M tokens.
Operational comparison
Create a simple comparison table for conceptual understanding (names anonymized). The values are representative:
Provider
TTFT (s)
Throughput (TPS)
Cost ($/1M tokens)
Provider A
0.32
544
0.18
Provider B
1.5
700
0.14
Provider C
0.27
313
0.16
Provider D
4.5
900
0.13
Provider A resembles Clarifai’s Reasoning Engine. Provider B emphasizes throughput at the expense of TTFT. Provider C may represent a hybrid player balancing both. Provider D shows that extremely high throughput can coincide with very poor TTFT and may only suit offline tasks.
Choosing the right provider
Startups building chatbots or assistants: Choose providers with low TTFT and moderate throughput; ensure you have instrumentation and the ability to tune prompts.
Batch pipelines: Select high‑throughput providers with good cost efficiency; ensure SLOs are still met.
Enterprises requiring flexibility: Evaluate whether the platform offers compute orchestration and local runners to deploy across clouds or on‑prem.
Regulated industries: Verify that the platform supports data residency and governance; Clarifai’s control center and fairness dashboards help with compliance.
Original Framework: Provider Fit Matrix
Plot TTFT on one axis and throughput on the other; overlay cost per million tokens and capability (e.g., local deployment, fairness tools). Use this matrix to decide which provider fits your persona (startup, enterprise, research) and workload (chatbot, batch generation, analytics).
Expert Insights
Independence matters: Benchmarks vary widely; ensure comparisons are done on the same model with the same prompts to make fair conclusions.
Clarifai differentiators: Clarifai’s compute orchestration and local runners enable on‑prem deployment and model portability; analytics dashboards provide real‑time TTFT and percentile latency monitoring.
Watch tail latencies: A provider with low average TTFT but high p99 latency may still yield poor user experience.
Quick Summary
What matters in benchmarks? TTFT, throughput, cost and deployment flexibility.
Which provider to choose? Match provider strengths to your persona and workload; for interactive apps, prioritize TTFT; for batch jobs, prioritize throughput and cost.
Caveats: Benchmarks are model‑specific; check data residency and compliance requirements.
Beyond Throughput – Introducing Goodput and Percentile Latencies
Why throughput isn’t enough
Throughput counts all tokens, regardless of how long they took to arrive. Goodput focuses on outputs that meet latency SLOs. A system may process 100 requests per second, but if only 30% meet the TTFT and TPOT targets, the goodput is effectively 30 r/s. The emerging consensus in 2025–2026 is that optimizing for goodput better aligns engineering with user satisfaction.
Defining and measuring goodput
Goodput is defined as the maximum sustained arrival rate at which a specified fraction of requests meet both TTFT and TPOT SLOs. For token‑level metrics, goodput can be expressed as the sum of outputs meeting SLO constraints divided by time. Emerging frameworks like smooth goodput further penalize prolonged user idle time and reward early completion.
To measure goodput:
Set SLO thresholds (e.g., TTFT <500 ms, TPOT <50 ms).
Instrument at fine granularity: log prefill completion, each token emission and request completion.
Compute the fraction of outputs meeting SLOs and divide by elapsed time.
Visualize percentile latencies (p50, p95, p99) to identify tail effects.
Clarifai’s analytics dashboard allows configuring alerts on p95/p99 latencies and goodput thresholds, making it easier to prevent SLO violations.
Goodput in the context of emerging architectures
Prefill‑decode disaggregation enables independent scaling of phases, improving both goodput and throughput. Advanced scheduling algorithms—length‑aware batching, SLO‑aware admission control and deadline‑aware scheduling—focus on maximizing goodput rather than raw throughput. Hardware‑software co‑design, such as specialized kernels for prefill and decode, further raises the ceiling.
Original Framework: Goodput Dashboard
A Goodput Dashboard should include:
Goodput over time vs. raw throughput.
Distribution of TTFT and TPOT to highlight tail latencies.
SLO compliance rate as a gauge (e.g., green above 95%, yellow 90–95%, red below 90%).
Phase utilization (prefill vs decode) to identify bottlenecks.
Per‑persona view: separate metrics for interactive vs batch clients.
Integrating this dashboard into your monitoring stack ensures engineering decisions remain aligned with user experience.
Expert Insights
Focus on user‑satisfying outputs: Research emphasises that goodput better captures user happiness than aggregate throughput.
Latency percentiles matter: High p99 latencies can cause a small subset of users to abandon sessions.
SLO‑aware algorithms: New scheduling approaches dynamically adjust batching and admission to maximize goodput.
Quick Summary
What is goodput? The rate of outputs meeting latency SLOs.
Why care? High throughput can mask slow outliers; goodput ensures user satisfaction.
How to measure? Instrument TTFT and TPOT, set SLOs, compute compliance, track percentile latencies and use dashboards.
Emerging Trends and Future Outlook (2026+)
Hardware, models and architectures
By 2026, new GPUs like NVIDIA’s H100 successor (H200/B200) offer higher memory bandwidth, enabling faster prefill and decode. Open‑source inference engines such as FlashInfer and PagedAttention reduce inter‑token latency by 30–70%. Research labs have shifted towards disaggregated architectures by default, and scheduling algorithms now adapt to workload patterns and network conditions. Models are more diverse: mixture‑of‑experts, multimodal and agentic models require flexible infrastructure.
Strategic implications
Hybrid deployment becomes the norm: Enterprises mix cloud, edge and on‑prem inference; Clarifai’s local runners support data sovereignty and low latency.
Configurable modes: Future systems may let users choose between Ultra Low TTFT and Maximum Throughput modes on the fly.
Goodput‑centric SLAs: Contracts will include goodput guarantees rather than raw TPS.
Responsible AI demands: Fairness dashboards, bias mitigation and audit logs become mandatory.
Original Framework: Future‑Readiness Checklist
To prepare for the evolving landscape:
Monitor hardware roadmaps: Plan upgrades based on memory bandwidth and local availability.
Adopt modular architectures: Ensure your serving stack can swap inference engines (e.g., vLLM, TensorRT‑LLM, FlashInfer) without rewrites.
Invest in observability: Track TTFT, TPOT, throughput, goodput and fairness metrics; use Clarifai’s analytics and fairness dashboards.
Plan for hybrid deployments: Use compute orchestration and local runners to run on cloud, edge and on‑prem simultaneously.
Stay up to date: Participate in open‑source communities; follow research on disaggregated serving and goodput algorithms.
Expert Insights
Disaggregation becomes default: By late 2025, almost all production‑grade frameworks adopted prefill‑decode disaggregation.
Latency improvements outpace Moore’s law: Serving systems improved more than 2× in 18 months, reducing both TTFT and cost.
Regulatory pressure rises: Data residency and AI‑specific regulation (e.g., EU AI Act) drive demand for local deployment and governance tools.
Quick Summary
What’s next? Faster GPUs, new inference engines (FlashInfer, PagedAttention), disaggregated serving, hybrid deployments and goodput‑centric SLAs.
How to prepare? Build modular, observable and compliant stacks using compute orchestration and local runners, and stay active in the community.
Key insight: Latency and throughput improvements will continue, but goodput and governance will define competitive advantage.
Frequently Asked Questions (FAQ)
What is TTFT and why does it matter?
TTFT stands for time‑to‑first‑token—the delay before the first output appears. It matters because it shapes user perception and trust. For interactive applications, aim for TTFT under 500 ms.
How is throughput different from goodput?
Throughput measures raw tokens or requests per second. Goodput counts only those outputs that meet latency SLOs, aligning better with user satisfaction.
Can I optimize both TTFT and throughput?
Yes, but there is a trade‑off. Use the Latency–Throughput Tuning Checklist: optimize prompts, choose smaller models, enable caching, adjust batch sizes and deploy near users. Monitor p95/p99 latencies and goodput to ensure one metric doesn’t sacrifice the other.
What is prefill‑decode disaggregation?
It’s an architecture that separates prompt ingestion (prefill) from token generation (decode), allowing independent scaling and reducing interference. Disaggregation has become the default for large‑scale serving and improves both TTFT and throughput.
How do Clarifai’s products help?
Clarifai’s compute orchestration spins up secure environments across clouds or on‑prem. Local runners let you deploy models near data sources, reducing network latency and meeting regulatory requirements. Model inference services support multiple models, with fairness dashboards for monitoring bias. Its analytics track TTFT, TPOT, TPS and goodput in real time.
By using frameworks like the Perception–Capacity Matrix and Latency–Throughput Tuning Checklist, focusing on goodput rather than raw throughput, and leveraging modern tools like Clarifai’s compute orchestration and local runners, teams can deliver AI experiences that feel instantaneous and scale efficiently into 2026 and beyond.
So you have an idea for the next “big” thing that will revolutionize the world. Maybe it struck you in the shower. Maybe it struck you in the wee hours of 3 in the morning when you’re half-awake and chatted with your cat. You drew something vaguely potato-wing-like on a napkin, and now you’re certain that it will shake up at least three markets. To the exhilarating and sometimes frightening world of hardware concept creation!
Hardware development is not a weekend hack-a-thon to build a new app. There are no quick patches or magical “undo” buttons when you find that your prototype’s battery roasts like a toaster oven on steroids. That is precisely why there are concept design services. It is the step that prevents your idea from becoming a costly paperweight.
Startups like to downplay how complicated this stage can be. There are drawings to figure out, user requirements to go over, materials to keep in mind, and prototyping techniques to schedule. Leave any of these behind, and you’ll have something lovely on Instagram but breakable in half when a toddler lays hands on it.
The best news is, you don’t have to do this on your own. Product design service firms are experts at taking goofy ideas and turning them into viable, manufacturable designs. Hiring professionals up front will protect you from unwarranted expense and torturous revisions. Places like Cad Crowd enable you to get in touch easily with seasoned product designers who both know how to be creative and also understand engineering. They’re your concept design safety net.
Here, we will take you through the basics of hardware concept design without blowing your mind with techno-jargon. We will also touch on why sketches matter more than you know, running tests on assumptions without spending a fortune, and what you can expect when working with design services companies. There will be some laughs, some cautionary tales, and plenty of real-world tips along the way. When you are finished, you will have a clear vision for taking your “potato with wings” and making it a polished product that has a legitimate chance at succeeding in the marketplace.
🚀 Table of contents
The thrill and terror of your first hardware idea
All entrepreneurs have experienced that shivery moment when a flash of inspiration hits you. Your mind leaps ahead to the media spotlights, the TED talk, and the yacht you will one day buy. But between your scribbling in the notebook and your first prototype, harsh reality will snap you back to attention with a large rubber glove.
Hardware does not play nice. While software may be a question of painting a virtual image, hardware is a question of sculpting marble. As soon as you take away too much, you can’t just hit “undo.” No, there is something about watching your “innovative” design splatter its initial drop test onto a cold concrete floor. It wobbles, flails, and then your brilliant idea is torn to smithereens like a disappointed LEGO set after a fit from a toddler.
This is where concept design saves your sanity. It has nothing to do with making pretty pictures. Concept design experts ask difficult questions before you invest hard money. It forces you to see options, consider functionality, and consider manufacturability. Good product design services companies will even shoot holes in your plan to prevent it from falling apart literally.
Hardware concept design is really the translation office from your imagination to the factory floor. What you are doing here is taking your idea and making it something that can be translated, tested, and ultimately made.
It involves sketches, renderings, rough models, and lots of “what if” talking. Unlike industrial design, which is deeply interested in form and appearance, or engineering design, which drills into technical detail, concept design is the playpen where art and science meet.
Consider a Venn diagram where one of the circles is “Looks Cool,” and the other is “Actually Works.” Concept design is the lovely overlap. It is why your shiny toy won’t need duct tape to function.
Good concept design is also narrative. A business-grade rendering or mockup tells investors, fellow colleagues, and potential clients, “This exists. This is happening.” And when you use a service like Cad Crowd to source designers, you are not just hiring a pair of hands. You are getting access to people who understand how to take your idea and make them understand clearly without your needing to defend your napkin scribble in a boardroom.
Turning brain sparks into tangible plans
Do not rush off to your CAD software or 3D printing design service just yet. Step back and ask yourself: Does anyone actually need this thing? Founders are prone to falling in love with what they’ve created, only to discover later that no one else wants it either.
Begin with market research. It does not need to be a grueling spreadsheet endurance test. Interview prospective users. Blog. Observe what people grumble about in criticisms of products that already exist. If your device resolves a genuine pain point, you are already ahead of half the startups in the world.
Next, describe your product’s major functions. What is it solving for you? What sets it apart? Keep it brief. There is a temptation to pack it with every conceivable feature. Now your sleek smart toothbrush doubles as a music player, weather checker, and espresso machine. That is feature creep, and it is the bane of good hardware design.
Product design companies can help here, too. They’ve seen what works, what doesn’t, and what eats through a budget faster than “crowdfunding fiasco.” A company you find through Cad Crowd can help your idea flow into a tight, buildable idea without your laying out one dime on tooling.
Sketches, renderings, and rough models
Don’t underestimate the authority of a poor drawing. Some of the greatest products ever created were badly drawn. Perfection isn’t the goal here. Communication is the goal.
Start with pencil sketches. Even if your drawings in elementary school were better, you can still mark principal shapes and functions. Once you have a number of promising leads, go to digital media like CAD.
They enable you to experiment with proportions, dimensions, and mechanical components more accurately.
Your professional design team can bring it to life. They make it possible for stakeholders to see your product as real. Photo-realistic images are something that product design experts can do well. These can be used to entice investors or test consumer appeal on social media. Platforms like Cad Crowd introduce you to designers who can turn your gadget into a million-dollar product before manufacturing a prototype.
If you’re the do-it-yourself type, you can make crude models using foam, cardboard, or even clay. Low-cost models allow you to try out size and ergonomics without jeopardizing expensive materials. You may find that your hand-held device is child-proof but painful for an adult. Worse to find that out than after a complete production run.
Prototyping without burning your wallet
At some point, your sketches and CAD models must leave the virtual world. That is where prototyping fits in. It is like the ugly teenager phase of your product. It is not yet attractive, but it is growing really fast. Rapid prototyping is within surprisingly easy reach. 3D printing lets you make physical models rapidly and inexpensively. You can try out shapes, fit, and even primitive functions without selling your kidney for the price of production. Foam models are another inexpensive way to check ergonomics. They are the action figure figurine form of your product: inexpensive, small, and surprisingly enlightening.
CNC machining costs more but produces more accurate and durable prototypes. It’s convenient if you need to test-run mechanical pieces or stress areas. Regardless of what you choose to do, don’t fall into the thinking trap that your first prototype must be perfect. It is meant to fail where you did not expect it. That is what it is for.
It is at this point that most entrepreneurs fear, believing that defective prototypes will scare away investors or partners. Actually, the fact that you are working towards refining and testing your idea indicates that you are mature. Product design services companies, especially those you can access through Cad Crowd, can guide you through prototyping design services without resource wastage. They know what areas can be started with for trial purposes and what can be done on the next round.
Manufacturing and material considerations prior to pledge
Oh, materials. They are the unsung heroes and sometimes villains of hardware creation. Get it wrong, and your product cracks, warps, or costs more to make than you wanted. Get it right, and you can save money, increase longevity, and get your product to be more appealing.
Start by considering the environment in which your product will exist. Will it see moisture, heat, or abuse? A fashion plastic can be stylish-looking, but it could melt faster than an ice cream cone at the beach on a hot summer day if it’s exposed to high temperatures. Metals are strong but heavy, and some composites are light but stiff. There are always compromises with each.
Manufacturability is as much a consideration as appearance. Some materials are harder to machine, mold, or assemble. If your product requires exotic parts or impossible tolerances, factories will avoid you or charge you an outrageous sum. It is for this reason that hiring a product design services company in the beginning is a good idea. They can identify manufacturing nightmares in advance before you become infatuated with a design that cannot be produced in quantity.
Another thing to consider is sustainability. Consumers have become increasingly conscious of saving the environment (as we all should). By using recyclable materials on your product, you can attract consumers that promotes sustainability. Like the designers from Cad Crowd, they can help you and your engineering design firm find a balance between sustainability, quality, and price.
Designing with product design services firms
Working with a product design services firm is like leaving your baby with a babysitter for the first time. You worry that they will mess it up or, worse still, will attempt to make it “better” in a way you don’t like. But a good design partner will treat your conception carefully and react with proficiency that you can never reproduce by yourself.
Start by looking for prospective partners. Look for portfolios that match your product’s level of style and complexity. Check out reviews and testimonials. Don’t be afraid to ask for references or samples of similar projects.
In any aspect of life, communication is always the key. Always know what you want, especially regarding the project deadline and allocation of money. Ask for updates, even if it’s a drawing or scribbles. A professional designer can’t deliver your expected results if you can’t communicate clearly what you need.
Legal protection is also important. Use of non-disclosure agreements (NDAs) to protect your intellectual property. Most professional CAD design services will require it and will willingly sign.
Platforms like Cad Crowd make it easier for you to complete your team. They can connect you with screen-tested designers and engineers specializing in hardware concept design or CAD. Especially, if you require assistance in creating starting sketches, 3D modeling, or production-ready file preparation, you can find the perfect match for you without wasting months asking for recommendations.
Lastly, don’t forget cooperation is a two-way street. Hear criticism out. Experienced designers may propose changes that maximize usability, minimize cost, or ease manufacturing. While it stings to leave behind your original idea, the changes often are between a product that fails and a product that succeeds.
Common mistakes first-time founders make
Every new hardware startup founder has a horror story to tell. Some of them are funny in hindsight. Some of them are a nightmare. Listening to them can spare you the expensive mistakes.
One of the biggest mistakes in business is adding more features to a product. It all starts with a simple idea, let’s say you want to produce a cooler bottle that keeps drinks cold for longer hours. Next, you want to include a Bluetooth speaker, a cup warmer, and a built-in blender. Now, the final product is more expensive than the latest game console. This reminds you to keep your product simple and not add unnecessary features just because it’s in the latest trend.
Another common mistake is neglect of design for manufacturability services. You may create a beautiful product that looks wonderful, but is impossible to produce at an affordable price. Factories are not magic. If you specify super-tiny tolerances or unique parts never made before, expect stratospheric prices or pleasant rejection letters.
Forgetting to test is another classic repeat offender. You’re convinced your design is perfect, so you launch straight into manufacturing. And you discover that the battery lasts only twenty minutes, or the hinges collapse at slight pressure. Iteration is dull, but it’s much cheaper than recalling ten thousand dead units.
Finally, poor communication can swamp even a good project. When you are vague with your product design services company, you invite miscommunication. Specific instructions, attainable expectations, and regular feedback prevent frustration from both sides. Cad Crowd makes it a lot easier to locate communicative, talented designers, but you still need to speak up.
Iteration isn’t sexy, but this is where the magic happens. Every incredible piece of hardware you adore, from your favorite headphones to your appliances in the kitchen, had thousands of iterations before it went to market. Even Apple and Dyson, those industry giants, test and iterate furiously.
Each revision is picking up something new. Maybe your device is cumbersome to hold, or maybe one button’s placement is confusing for users. These small adjustments, repeated time and again, make a good product into a great product through prototype engineering services.
User testing is worth its weight in gold. Show your prototype to someone who has never seen it before and observe what they do. They will press the wrong button, flip it over, or use it in a way that you never dreamed. It’s an eye-opener, but it teaches you things that no spreadsheet ever could.
Never give up on failure. Treat every mistake as a veil for progress. The earlier your mistakes, the earlier your growth. Work with your product design services company to incorporate critiques, re-doing designs, and building improved prototypes. Cad Crowd’s network of designers can guide you through such loops efficiently and steer clear of wasteful setbacks.
Bonus tips for navigating the wild world of hardware design
Even with the best design, hardware development tends to surprise you. These are some other tricks to put in your regular bag and maybe avoid a surprise crash:
But packaging can shatter or create the first customer impression. Packaging design services decide shipping expenses, shelf life, and even safety. Involve your product design company in packaging decisions. Designers on sites like Cad Crowd can suggest functionalities that are affordable, aesthetic, and functional.
Don’t accept the first manufacturing offer at face value. Research alternatives like injection molding, vacuum forming, or die casting. Another option, usually, will be cheaper or improve the quality. A skilled designer will walk you through the pros and cons without getting bogged down with details.
Real-life situations
Just imagine the pet feeder; the first designs were stunning on paper. However, when the first batch of manufactured products arrived, the hinge on the lid failed after a few cycles. Customers posted videos of cats blissfully trashing the feeders like furry little engineers. The company recalled the product and re-engineered the hinge, an expensive lesson in testing in real-world environments.
And another founder wanted to create a wearable fitness tracker with a revolutionary clasp system. They skipped user testing because “everyone knows how to use a clasp.” But they didn’t. The clasp was so user-unfriendly that customers wore the tracker backwards or broke it attempting to buckle it. A single round of user testing could have avoided months of embarrassment and lost sales.
It’s because of anecdotes like these that iteration, open feedback, and collaboration with product development experts are so crucial. A good designer will not just provide you with neat files. They will burst bubbles in assumptions, suggest ways to make it better, and guide you around pitfalls that have swallowed up other founders whole.
It’s easy to take only care of the physical aspects of your product and neglect branding. Branding, however, is not merely logos and color schemes. Branding is what informs your design decisions from the start. Is your product sleek and futuristic? Friendly and playful? Outdoor-inspired and tough? These decisions determine everything from material selection to button shape.
Your designer can incorporate branding into the concept phase. If, for instance, your company is concerned about sustainability, that should guide your selection of materials and packaging. If your item is aimed at a high-end market, your concept design should be sophisticated and precise. Cad Crowd’s network has designers who know branding as part of the larger picture and will make sure your product and your brand are a natural fit.
Getting ready for manufacturing like a pro
Once you have already planned your product, you must prepare for large-scale production. This is the most underestimated phase for the first-time businessman. For them to create a successful product, they need detailed drawings, clear specifications, and clear illustrations. That’s why sloppy documentation and management can result in a very expensive failure.
Work with your product design services firm to create manufacturing-ready documents. Double-check every measurement, every tolerance, and every material specification. Don’t rely on the manufacturing design firm to “figure it out.” They will produce exactly what you give them, which can mean producing 10,000 pieces with a defect that could have been identified early on.
You ought also to think about production locations. Home-country production can be easier in terms of communication and quality control, but foreign production can be less expensive, but needs to have great screening and perhaps longer lead times. Think in terms of shipping, tariffs, and variations in communication styles. A seasoned global manufacturing designer will make this easier to do.
Holding sanity together through the highs and lows
Hardware construction is a series of rollercoasters. There will be a time when you will wonder why in the world you ever thought that your plan was a good idea, but that’s fine. Remember to surround yourself with positive peers, mentors, or even online communities of fellow founders. Share your stories, and sometimes, a word of encouragement here and there from someone who has walked in your footsteps can make all the difference.
Failures are redirections. These are your long-term motivations. Keep in mind that you’re making something real, that people can hold, use, and keep close to their hearts. That’s worth the headaches. Knowing you’re leaving a legacy.
You now have information, game plans, and a little bit of sound advice. But information does not build a product. Action does. Start sketching. Research your market. Get in touch with a product design services company. Sites like Cad Crowd wait in the wings with a group of good designers who can transform your “someday” idea into a real, producible product.
Don’t wait until your concept is perfected because it won’t be. Perfection is the progress killer. The sooner you get your idea in motion, the sooner you can learn, adapt, and build something amazing. The world doesn’t need another napkin drawing that has been left behind. It needs your idea, refined, experimented with, and ready to take over the world. Request a quote today.
MacKenzie Brown is the founder and CEO of Cad Crowd. With over 18 years of experience in launching and scaling platforms specializing in CAD services, product design, manufacturing, hardware, and software development, MacKenzie is a recognized authority in the engineering industry. Under his leadership, Cad Crowd serves esteemed clients like NASA, JPL, the U.S. Navy, and Fortune 500 companies, empowering innovators with access to high-quality design and engineering talent.
I’m pretty horrified that Pokémon is now 30 years old, so I’ve decided to fight the existential dread by playing N64 all weekend. Specifically, I’m planning on playing Pikachu’s 3D debut on an actual Nintendo 64, potentially with some friends if they’re up for a Pokémon Stadium party, so I thought I’d share the setups I’m planning to use.
When it comes to using the best retro consoles, I’m constantly swapping in and out systems and gadgets. To celebrate Pokémon’s 30th Anniversary, it feels only right that I use the special edition Pikachu N64 console from the year 2000, but you don’t need a big electric mouse on the front of your console to replicate my setup. Basically, my ensemble should act as inspiration for revising arguably one of the greatest creature-clashing party game hybrids of all time.
What I will say is that there isn’t a wrong way to play Pokémon Stadium in 2026, so long as you’re having fun. I’ll be detailing how I’m still using an actual N64, but there’s no shame in using one of the best retro handhelds to revisit the turn-based battler on the go. If you’d rather keep things authentic for the big 30th Anniversary, though, I’ve got you covered with retro controller options, ways to plug into modern screens, and more.
Pika the right screen and cables
(Image credit: Phil Hayton)
If you’re up for some Pokémon Stadium party planning, you’ll want to start by getting the N64 connected to a screen. Checking out how to connect your retro console to a modern TV or monitor is a nice place to start, but if you’re like me, you might be using an old CRT TV from 30 years ago.
Those of you who have managed to keep onto or find a CRT might already be familiar with how to plug in an N64. But, if you’re just starting out, you’ll want to effectively hook up using either an RCA composite (the red, blue, and yellow jacks) or S-Video if you’re feeling fancy. It’s worth noting that if you are lucky enough to own a PAL Pikachu N64, the latter isn’t going to work with the console since it doesn’t support the output, and I’ve noticed some awful interference in NTSC versions that do retain the feature.
Depending on how old your TV is, there’s a chance you won’t find those aforementioned RCA inputs. If that’s the case, I’d strongly recommend looking for another screen rather than using an RF cable to tune into your set. You can technically do this and get playing, but the visuals will look especially scrappy in 2026 thanks to interference from other modern signals.
(Image credit: Future / Phil Hayton)
As appropriate as the Pikachu N64 is, you’ll actually have a smoother time using the OG grey N64. That especially applies if you’re using modern displays, as it means you can HDMI mod your N64 and hook up straight to your TV, or even use an adapter to convert a higher quality S-Video signal. The latter will set you back around $24.99 at Amazon, and while you’ll get better results if you use a dedicated upscaler like the OSSC (like this one for $111), opting for the straight converter is the quickest way to hook up.
Weekly digests, tales from the communities you love, and more
You might have noticed that I’m personally using a Sony PVM CRT and a Pikachu N64, and while it looks extra crispy and vibrant, you absolutely don’t have to go down that route. In truth, it’s not the only way I’ve been playing Pokémon Stadium lately, as I’ve also jammed my cart into the 4K Analogue 3D and my modded console.
Image 1 of 1
(Image credit: Phil Hayton)
If you can afford the $269.99 FPGA remake, it’s going to look far nicer on large screens, thanks to the extremely accurate CRT filters. My DIY version also looks nicer if you pair it with a scanline generator for some needed pixel separation, but the ultimate goal really is to just see some Pokémans successfully show up on your display of choice.
8Bitdo N64 controller, I choose you!
(Image credit: Phil Hayton)
The next order of business is plugging a gamepad into Ninty’s console. I mean, you’ll need it if you’re going to frantically feed Lickitung in Sushi Go-Round, and while you’ll be able to feed the lil guys using your original gamepad, I reckon most of you will want to use the 8Bitdo N64 controller since it blends the authentic button layout with something that feels like a modern pad.
I reckon that if you are planning a four-player Pokémon Stadium party, grabbing the multiplayer Blueretro N64 adapter could be the right call. The one I use is around $15 at Aliexpress, and it means you can use both the 8BitDo pad and another other controller you happen to have lying around, including the DualSense and other Bluetooth options. That sure as hell beats being handed a crummy “little brother” third-party controller back in the ’90s, and it’s a cheaper solution than picking up four specific gamepads.
Of course, if you do opt for a modern gamepad, you’ll be missing the rumble and memory card slot at the back. This isn’t really an issue if you’re just looking to save games since the Blueretro adapter provides some storage while enabling vibration on synced controllers, but it does mean you won’t be able to use the Transfer Pak for importing your Game Boy party or playing your cart on the big screen. If you’d rather retain that, your best bet is to look at something like the Hyperkin Admiral for $19.99 that is effectively a clone of the original design.
(Image credit: Phil Hayton)
Actually diving back into Pokémon Stadium, even on an original N64, is pretty simple even in 2026. So long as you sort out your screen connections and pick out your preferred controller, you should be good to go. If you are looking for extras, though, there are bits and bobs that can add something extra to your old console.
The first is handy if you’re into exploring Pokémon Studium ROM hacks, as you can do that using a Summercart 64 flash cart. I’ve previously highlighted how you can use the gadget to play Zelda fan games, and while it’s slightly pricey at $82.99, it’ll open a door to a whole world of custom adventures, custom mods, and even unreleased versions of classics.
If you have an old N64 in your dusty attic, or happen to come across a console that’s perhaps slightly grimy, it’s also worth grabbing a cartridge slot cleaner. No, you’re not really supposed to blow into the slot until it eventually works, and while I’ll not judge you as much as other enthusiasts for doing that, you’ll either want to buy a dedicated “1Upcard” for $24.99 or something similar if you can find something cheaper. If that sounds steep, you could instead make your own with a soft glasses cloth, some thin cardboard, and potentially a splash of isopropyl alcohol, but it is safer just sticking with something designed for the job.
Hopefully, my N64 setup nonsense will inspire some of you to have a Pokémon Stadium party of your own over the weekend. If anything, I’m really hoping that my words simply break down any imaginary barrier to using the original console as the series turns 30, but as I said before, I won’t judge you if you choose to just use a gaming handheld or even the Switch Online version.
Friday afternoon, just as this interview was getting underway, a news alert flashed across my computer screen: the Trump administration was severing ties with Anthropic, the San Francisco AI company founded in 2021 by Dario Amodei. Defense Secretary Pete Hegseth soon after invoked a national security law to blacklist the company from doing business with the Pentagon after Amodei refused to allow Anthropic’s tech to be used for mass surveillance of U.S. citizens or for autonomous armed drones that could select and kill targets without human input.
It was a jaw-dropping sequence of events. Anthropic stands to lose a contract worth up to $200 million and could be barred from working with other defense contractors after President Trump posted on Truth Social directing every federal agency to “immediately cease all use of Anthropic technology.” (Anthropic has since said it will challenge the Pentagon in court.)
Max Tegmark has spent the better part of a decade warning that the race to build ever-more-powerful AI systems is outpacing the world’s ability to govern them. The MIT physicist founded the Future of Life Institute in 2014 and in 2023 helped organize an open letter — ultimately signed by more than 33,000 people, including Elon Musk — calling for a pause in advanced AI development.
His view of the Anthropic crisis is unsparing: the company, like its rivals, has sown the seeds of its own predicament. Tegmark’s argument doesn’t begin with the Pentagon but with a decision made years earlier — a choice, shared across the industry, to resist regulation. Anthropic, OpenAI, Google DeepMind and others have long promised to govern themselves responsibly. Anthropic this week even dropped the central tenet of its own safety pledge — its promise not to release increasingly powerful AI systems until the company was confident they wouldn’t cause harm.
Now, in the absence of rules, there’s not a lot to protect these players, says Tegmark. Here’s more from that interview, edited for length and clarity. You can hear the full conversation this coming week on TechCrunch’s StrictlyVC Download podcast.
When you saw this news just now about Anthropic, what was your first reaction?
The road to hell is paved with good intentions. It’s so interesting to think back a decade ago, when people were so excited about how we were going to make artificial intelligence to cure cancer, to grow the prosperity in America and make America strong. And here we are now where the U.S. government is pissed off at this company for not wanting AI to be used for domestic mass surveillance of Americans, and also not wanting to have killer robots that can autonomously — without any human input at all — decide who gets killed.
Techcrunch event
San Francisco, CA | October 13-15, 2026
Anthropic has staked its entire identity on being a safety-first AI company, and yet it was collaborating with defense and intelligence agencies [dating back to at least 2024]. Do you think that’s at all contradictory?
It is contradictory. If I can give a little cynical take on this — yes, Anthropic has been very good at marketing themselves as all about safety. But if you actually look at the facts rather than the claims, what you see is that Anthropic, OpenAI, Google DeepMind and xAI have all talked a lot about how they care about safety. None of them has come out supporting binding safety regulation the way we have in other industries. And all four of these companies have now broken their own promises. First we had Google — this big slogan, ‘Don’t be evil.’ Then they dropped that. Then they dropped another longer commitment that basically said they promised not to do harm with AI. They dropped that so they could sell AI for surveillance and weapons. OpenAI just dropped the word safety from their mission statement. xAI shut down their whole safety team. And now Anthropic, earlier in the week, dropped their most important safety commitment — the promise not to release powerful AI systems until they were sure they weren’t going to cause harm.
How did companies that made such prominent safety commitments end up in this position?
All of these companies, especially OpenAI and Google DeepMind but to some extent also Anthropic, have persistently lobbied against regulation of AI, saying, ‘Just trust us, we’re going to regulate ourselves.’ And they’ve successfully lobbied. So we right now have less regulation on AI systems in America than on sandwiches. You know, if you want to open a sandwich shop and the health inspector finds 15 rats in the kitchen, he won’t let you sell any sandwiches until you fix it. But if you say, ‘Don’t worry, I’m not going to sell sandwiches, I’m going to sell AI girlfriends for 11-year-olds, and they’ve been linked to suicides in the past, and then I’m going to release something called superintelligence which might overthrow the U.S. government, but I have a good feeling about mine’ — the inspector has to say, ‘Fine, go ahead, just don’t sell sandwiches.’
There’s food safety regulation and no AI regulation.
And this, I feel, all of these companies really share the blame for. Because if they had taken all these promises that they made back in the day for how they were going to be so safe and goody-goody, and gotten together, and then gone to the government and said, ‘Please take our voluntary commitments and turn them into U.S. law that binds even our most sloppy competitors’ — this would have happened. Instead, we’re in a complete regulatory vacuum. And we know what happens when there’s a complete corporate amnesty: you get thalidomide, you get tobacco companies pushing cigarettes on kids, you get asbestos causing lung cancer. So it’s sort of ironic that their own resistance to having laws saying what’s okay and not okay to do with AI is now coming back and biting them.
There is no law right now against building AI to kill Americans, so the government can just suddenly ask for it. If the companies themselves had earlier come out and said, ‘We want this law,’ they wouldn’t be in this pickle. They really shot themselves in the foot.
The companies’ counter-argument is always the race with China — if American companies don’t do such and such, Beijing will. Does that argument hold?
Let’s analyze that. The most common talking point from the lobbyists for the AI companies — they’re now better funded and more numerous than the lobbyists from the fossil fuel industry, the pharma industry and the military-industrial complex combined — is that whenever anyone proposes any kind of regulation, they say, ‘But China.’ So let’s look at that. China is in the process of banning AI girlfriends outright. Not just age limits — they’re looking at banning all anthropomorphic AI. Why? Not because they want to please America but because they feel this is screwing up Chinese youth and making China weak. Obviously, it’s making American youth weak, too.
And when people say we have to race to build superintelligence so we can win against China — when we don’t actually know how to control superintelligence, so that the default outcome is that humanity loses control of Earth to alien machines — guess what? The Chinese Communist Party really likes control. Who in their right mind thinks that Xi Jinping is going to tolerate some Chinese AI company building something that overthrows the Chinese government? No way. It’s clearly really bad for the American government too if it gets overthrown in a coup by the first American company to build superintelligence. This is a national security threat.
That’s compelling framing — superintelligence as a national security threat, not an asset. Do you see that view gaining traction in Washington?
I think if people in the national security community listen to Dario Amodei describe his vision — he’s given a famous speech where he says we’ll soon have a country of geniuses in a data center — they might start thinking: ‘Wait, did Dario just use the word country? Maybe I should put that country of geniuses in a data center on the same threat list I’m keeping tabs on, because that sounds threatening to the U.S. government.’ And I think fairly soon, enough people in the U.S. national security community are going to realize that uncontrollable superintelligence is a threat, not a tool. This is totally analogous to the Cold War. There was a race for dominance — economic and military — against the Soviet Union. We Americans won that one without ever engaging in the second race, which was to see who could put the most nuclear craters in the other superpower. People realized that was just suicide. No one wins. The same logic applies here.
What does all of this mean for the pace of AI development more broadly? And how close do you think we are to the systems you’re describing?
Six years ago, almost every expert in AI I knew predicted we were decades away from having AI that could master language and knowledge at human level — maybe 2040, maybe 2050. They were all wrong, because we already have that now. We’ve seen AI progress quite rapidly from high school level to college level to PhD level to university professor level in some areas. Last year, AI won the gold medal at the International Mathematics Olympiad, which is about as difficult as human tasks get. I wrote a paper together with Yoshua Bengio, Dan Hendrycks, and other top AI researchers just a few months ago giving a rigorous definition of AGI. According to this, GPT-4 was 27% of the way there. GPT-5 was 57% of the way there. So we’re not there yet, but going from 27% to 57% that quickly suggests it might not be that long.
When I lectured to my students yesterday at MIT, I told them that even if it takes four years, that means when they graduate, they might not be able to get any jobs anymore. It’s certainly not too soon to start preparing for it.
Anthropic is now blacklisted. I’m curious to see what happens next — will the other AI giants stand with it and say, ‘We won’t do this either?’ Or does someone like xAI raise their hand and say, ‘Anthropic didn’t want that contract, we’ll take it’?[Editor’s note: Hours after the interview, OpenAI announced its own deal with the Pentagon.]
Last night, Sam Altman came out and said he stands with Anthropic and has the same red lines. I admire him for the courage of saying that. Google, as of when we started this interview, had said nothing. If they just stay quiet, I think that’s incredibly embarrassing for them as a company, and a lot of their staff will feel the same. We haven’t heard anything from xAI yet either. So it’ll be interesting to see. Basically, there’s this moment where everybody has to show their true colors.
Is there a version of this where the outcome is actually good?
Yes, and this is why I’m actually optimistic in a strange way. There’s such an obvious alternative here. If we just start treating AI companies like any other companies — drop the corporate amnesty — they would clearly have to do something like a clinical trial before they released something this powerful, and demonstrate to independent experts that they know how to control it. Then we get a golden age with all the good stuff from AI, without the existential angst. That’s not the path we’re on right now. But it could be.
Whether you’re an adult or a child, going to the dentist can be a scary thing. You never know if you have a cavity or not and if you do, it’s painful to have it taken care of…not to mention expensive, sometimes expensive even if you have insurance.
It’s never a fun or pleasant visit, although dentists’ offices really try to make it as pleasant and pain-free as possible. They really do such a great job at trying to calm down our fears and put us at ease. ❤️ Still, it’s a scary thing, and for a kid, it’s even MORE scary.
Someone putting a tool in your mouth seems evasive, like a loss of control.
Will they hurt me?
Can I trust them?
Will it be painful?
Lots of worry and fret spin around in a child’s mind and so today, I wanted to help ease those fears by creating a super cute kawaii Hello Dentist Coloring Page Free Printable.
This adorable little PDF printable CAN absolutely help settle their nerves and bring a positive note to the whole dental experience and that’s my goal for you today. 🥰
Whether you’re prepping for a big dentist appointment moment or just looking for preschool dental health printables to keep them busy, bringing a dental coloring page printable to your dentist will really help give them something to focus on while they wait.
If you’re a parent or a teacher heading into dental health month activities, this children’s dental health coloring page is a total lifesaver in teaching children about dental hygiene.
If you work in a dentist’s office, grab this printable, make some copies, and put it in your lobby with a cup of crayons to help kiddos who come in and need something to do.
If you’re a parent, grab a couple teeth coloring page printables in your purse as you head out with your kids and let them color while they wait.
The cute little happy kawaii graphics are sure to make them feel more at ease and show them that it’s okay!!! ❤️
Kawaii Dentist Coloring Page For Kids
From the tiny molar characters to the little ‘fresh’ mouthwash bottles, every inch of this dental hygiene coloring sheet is designed to make oral dental health time feel like a happy adventure instead of a chore, so you can keep up on your cavity prevention plan with your little ones! 🥰
This cute dentist coloring page printable is packed with adorable dental doodles like happy teeth, a tooth with braces, toothpaste, mouthwash, a toothbrush, a dental mirror, apples, little hearts and stars, and 2-minute brushing timers, all in a clean black and white style that’s so satisfying to color.
It’s the perfect healthy teeth coloring page, no matter what age your children are, from toddlers to teens and everything in between!
You wanna get your kids to feel safe and to ease nerves. Using a pediatric dentist coloring page or other preschool dental health printables is such a great way to prep ’em! Here’s some great ideas that’ll help!
Dentist Role Play Game
Before the appointment, play “dentist” at home for two minutes.
Let your child be the dentist first, then you switch.
Count teeth, pretend to polish, and practice opening wide while staying relaxed and silly.
You can even show them this dentist tools coloring page to make toddler teeth coloring practice feel like a game. This will help them be at ease at the actual dentist office. ❤️
After the Dentist Reward
Plan a simple reward before you go so everyone knows what’s coming.
Offering a dental coloring page printable or a fun oral health coloring page as a little prize is such a great add-on to a happy experience. Here are some more fun ideas:
Park time
Library trip
Choosing dinner
Gift: a small toy, a new pencil, or even this cute brushing teeth printable
Picking the family movie
It’s definitely not an easy task for parents to go through (or for the kids), but helping them both before AND after the appointment will really solidify that going to the dentist is a normal, routine part of life that we all have to do and that it’s okay to be scared, but we need to be brave and take care of our health properly.
What sets your printables apart?
• Other printables are either functional OR cute. I bridge the gap between the two. Since I’m a single mom and don’t have time to mess around and am a kid at heart, I love something that actually propels me forward and gets the job done quickly, while still having that super cute vibe I’m known for. • I’m a graphic designer FIRST. This means my printables, whether free or paid for, will always be of the highest quality and print crisp and clear on any printer. • I love what I do and I do the work myself. I’m not hiring these out. This is my life, my hobby, my passion, my heart. • My fans are obsessed with my work. To date, I’ve made over $23M on printables alone (organically) in the past 8 years.
Is this printable free to download?
Yes! Except for some large binders, all printables on the blog are 100% free for personal use. I make these because I genuinely want you to have something cute and helpful right now without having to spend a dime. You deserve a life that feels more peaceful, more organized, and more you, and if a simple printable can make your day feel lighter, I am so happy to give that to you.
All printables are for personal use only.
Printables, graphics, backgrounds, SVGs, fonts, and design elements on this site are for personal use only. You may not alter them or redistribute them for free or monetary gain without written consent from the author, Sarah Titus. If you want to share these resources with others, please share the link to the blog post. Teachers & Churches: You may share these with your classes, colleagues, and congregation as long as they are not edited in any way.
Can I use these printables for my classroom or church group?
Absolutely! You’re welcome to print as many copies as you need for your own students or church groups. I only ask that you do not edit them in any way (leaving the full copyright line in tact) or sell them or host the digital files on other websites. Please always link back to this post to share with others.
What’s the best way to print these high-quality printables?
For best results, I recommend using standard 8.5″x11″ white paper. Since these are professionally designed as high-resolution PDFs, they’ll stay crisp and clear on any printer. To print, simply download the PDF to your device and open your printer settings. Select “Fit to Page” to ensure the printable looks perfect.
TL;DR: Watch the 2026 MotoGP World Championship for free on ServusTV. Access this free streaming platform from anywhere in the world with ExpressVPN.
Are you looking for thrilling and unpredictable racing? No we’re not talking about F1, unless you like regular pit stops, safety cars, and the same driver winning almost every week. We’re talking about a form of two-wheel racing where anything could happen.
MotoGP is where the real action can be found. Every week you see the best riders in the world go wheel to wheel with their rivals, throwing their bikes into corners with absolutely no sense of self preservation. It’s an awesome spectacle, and it doesn’t need to cost you anything to watch.
Want to watch the 2026 MotoGP World Championship for free from anywhere in the world? We have all the information you need.
What is MotoGP?
MotoGP is the oldest established motorsport world championship, with the inaugural season taking place in 1949. Races last approximately 45 minutes, without pitting for fuel or fresh tires.
The likes of Ducati, Honda, and Yamaha do battle for the constructors’ championship. Marc Márquez is the defending champion.
When is the 2026 MotoGP World Championship?
The 2026 MotoGP World Championship features 22 races held all over the world between March and November:
Mashable Top Stories
Thailand GP — March 1
Brazil GP — March 22
USA GP — March 29
Qatar GP — April 12
Spain GP — April 26
France GP — May 10
Catalonia GP — May 17
Italy GP — May 31
Hungary GP — June 7
Czechia GP — June 21
Netherlands GP — June 28
Germany GP — July 12
Great Britain GP — Aug. 9
Aragon GP — Aug. 30
San Marino GP — Sept. 13
Austria GP — Sept. 20
Japan GP — Oct. 4
Indonesian GP — Oct. 11
Australia GP — Oct. 25
Malaysia GP — Nov. 1
Portugal GP — Nov. 15
Valencia GP — Nov. 22
The good news for fans is that it’s possible to live stream the 2026 MotoGP World Championship without spending anything.
How to watch the 2026 MotoGP World Championship for free
The 2026 MotoGP World Championship is available to live stream for free on ServusTV.
ServusTV is geo-restricted to Austria, but anyone can access this free streaming platform with a VPN. These tools can hide your real IP address (digital location) and connect you to a secure server in Austria, meaning you can access ServusTV from anywhere in the world.
Unblock ServusTV by following this quick and easy process:
The best VPNs for streaming are not free, but they do tend to offer money-back guarantees. By using these money-back guarantees, you can watch MotoGP live streams without fully committing with your cash. This obviously isn’t a long-term solution, but it does mean you can watch select races without actually spending anything.
If you want to retain permanent access to free streaming services from around the world, you will need a subscription. Fortunately, the best VPN for live streaming sport is on sale for a limited time.
What is the best VPN for ServusTV?
ExpressVPN is tough to beat when it comes to streaming live sport, for a number of reasons:
Servers in 105 countries including Austria
Easy-to-use app available on all major devices including iPhone, Android, Windows, Mac, and more
Strict no-logging policy so your data is secure
Impressive streaming speeds without buffering
Up to 10 simultaneous connections
30-day money-back guarantee
A one-year subscription to ExpressVPN is on sale for $99.95 and includes an extra three months for free — 49% off for a limited time. This plan also includes a year of free unlimited cloud backup and a generous 30-day money-back guarantee. Alternatively, you can get a one-month plan for just $12.99 (with money-back guarantee).
Live stream the 2026 MotoGP World Championship for free with ExpressVPN.
Who would’ve guessed we’d get to play the original Banjo-Kazooie on a handheld with just a D-pad in 2026. HyperMegaTech!’s latest release is a collaboration with Rare Ltd., the legendary game developer known for the Banjo-Kazooie franchise and, more recently, Sea of Thieves, called the Super Pocket Rare Edition.
The vertical handheld features 14 classics from the British developer, including two Battletoads titles, Conker’s Pocket Tales and many more. While most of the games were released on 8- or 16-bit consoles, Banjo-Kazooie will be the headliner since it was originally released on the Nintendo 64. It may sound weird to control Banjo and Kazooie with a D-pad, but HyperMegaTech! assured that the game has been enhanced and optimized specifically for the Super Pocket handheld.
Since HyperMegaTech! and Evercade share Blaze Entertainment as a parent company, that means the Rare Edition handheld will be compatible with Evercade cartridges. Once you’re done with the 14 included games, you can expand your Super Pocket’s library with cartridges that feature collections from Taito, NeoGeo or Atari. HyperMegaTech! said the Rare Edition handheld will be available for $69.99 in June 2026, but has already opened preorders.
This is an exclusive column featuring expert analysts from International Data Corporation (IDC), who provide insights into the latest products, news, and more.
2026 marks the 20th anniversary of MWC in Barcelona, and instead of a big fat cake to celebrate, the mobile industry is facing one of the most unprecedented challenges in its history.
If you’re heading to Barcelona for Mobile World Congress 2026 expecting new Android phones and flashy demos, you’ll certainly not be disappointed. However, the real story runs deeper.
AI is becoming core to Android devices, foldables, and maturing into credible premium contenders, satellite connectivity is edging toward mainstream adoption, and a global memory crunch is reshaping pricing, upgrade cycles, and the broader industry landscape.
As an IDC market analyst, these are the themes I’ll be watching closely at MWC this year.
AI on smartphones: From feature to foundation
(Image credit: Nicholas Sutrich / Android Central)
In 2024 and 2025, the smartphone industry was dominated by an experimental “AI awe” phase. Most devices announced were labelled AI-enabled, but despite the excitement, most felt experimental. In 2026, the defining shift will be the transition of generative AI from a “cool” feature to the device’s foundational architecture.
IDC forecasts that GenAI-enabled smartphones will account for over 37% of total shipments in 2026, representing a market value of $433 billion. The segment is expected to grow at an average annual rate of 32% over the next five years. Critically, the democratization of AI is finally reaching mid-range devices, enabling personalized, proactive assistants to run locally rather than relying on the cloud.
AI becomes foundational. On Android, that shift is particularly visible, as chipmakers such as Qualcomm and MediaTek, alongside Google’s own platform strategy, are expanding on on-device AI performance. More GenAI tasks are now processed locally rather than in the cloud.
Get the latest news from Android Central, your trusted companion in the world of Android
However, the bigger change is architectural. AI is embedded into the OS itself, in app permissions, background services, and system-level workflows. It is no longer a standalone feature, but a structural component of the user experience. This will be a key message from most OEMs as they unveil their latest flagships.
From a market perspective, AI has become one of the primary drivers for sustaining premium pricing in a mature smartphone market.
The rise of agentic AI: When your Phone acts on your behalf
(Image credit: Honor)
AI agents will dominate conversations at MWC. OEMs are expected to showcase on-device AI agents capable of autonomously executing multi-step tasks such as planning travel, coordinating schedules, and managing smart home environments. This will offer an enhanced user experience when compared to simply responding to prompts.
The technical foundation is already visible on Apple Intelligence or Samsung Galaxy AI. On Android, this will mean deeper Google Gemini integrations, smoother cross-app coordination, persistent contextual memory, and greater use of powerful NPUs in flagship chipsets.
Consumers will benefit from faster, more seamless task execution, while smartphone makers gain ecosystem stickiness.
However, agentic AI raises a strategic question: trust. When devices begin to act on the user’s behalf, privacy, transparency, and control become critical. Android’s openness remains a competitive advantage, but it also introduces complexity in delivering consistent and secure AI experiences across multiple vendors.
Foldables still account for a single-digit share of total smartphone shipments (1.6% in 2025, according to IDC), but within the Android ecosystem, they play an outsized strategic role, accounting for 15% of the premium segment.
Samsung, Honor, Xiaomi, and others continue refining fold-type foldables as productivity-first devices rather than novelty hardware.
At MWC, we expect thinner designs, improved hinge durability, and reduced display creasing to narrow the gap with traditional bar-type smartphones.
From a market standpoint, foldables are less about volume and more about defending the premium tier. In a world where smartphone innovation feels incremental, foldables offer differentiation and margin protection.
The real question isn’t whether foldables will dominate shipments in 2026 — they will not. It’s whether they will become the growth engine at the high end of the market.
The memory crunch: Pricing pressure increases
(Image credit: Future)
While AI dominates the headlines, supply chain dynamics are becoming more influential.
Memory manufacturers are prioritising high-bandwidth memory (HBM) for AI data centres, tightening supply for the DRAM and NAND components used in smartphones. This reallocation is placing upward pressure on component costs across consumer devices.
For Android buyers, this may mean fewer aggressive mid-range price cuts and more cautious storage configurations at entry levels. For OEMs, the challenge is balancing AI-driven performance requirements, which demand more memory, with consumer sensitivity to rising prices.
For years, the industry has relied on predictable declines in component costs to sustain innovation without inflating retail prices. That assumption is now under strain, and it may influence upgrade cycles more than any single feature announcement.
IDC expects the memory supply challenges to persist throughout 2026 and likely well into 2027. There is no scenario in our current model in which memory pricing returns to 2025 levels within the forecast period
The downstream consequences of the crisis are becoming clearer and will reshape competitive dynamics in the PC and smartphone markets, and will be at the top of every conversation in Barcelona next week.
Power innovation: Charging and battery chemistry
(Image credit: Nicholas Sutrich / Android Central)
Battery technology is quietly becoming one of the most meaningful enablers of next-generation Android devices.
Ultra-fast charging, with demonstrations approaching 300W in controlled environments, shifts the narrative from “all-day battery” to “minutes of charging.” This could change user behaviour from overnight charging to rapid top-ups throughout the day. The real test, however, will be long-term battery health and thermal management.
At the same time, silicon-carbon anode batteries are increasing energy density compared to traditional graphite lithium-ion cells. This allows manufacturers to pack larger capacities into the same footprint or design thinner devices without sacrificing endurance.
As on-device AI workloads and high-refresh displays push power consumption higher, improvements in battery chemistry become strategically essential rather than incremental. At MWC, we expect OEMs to frame battery and charging not as secondary specifications, but as core enablers of AI performance.
Satellite connectivity: Expanding the coverage map
(Image credit: Derrek Lee / Android Central)
Satellite connectivity is evolving beyond emergency-only use cases. Discussions at MWC are expected to focus on non-terrestrial networks and expanded messaging capabilities, alongside deeper partnerships between mobile operators and satellite providers.
For Android users, satellite integration promises extended coverage in remote areas and improved resilience during natural disasters. The technology itself is largely proven. The next phase will depend on commercial models, carrier integration, and pricing structures that make satellite access practical rather than purely promotional.
Smart glasses
(Image credit: Nicholas Sutrich / Android Central)
AI-powered smart glasses are emerging as a complementary interface to the smartphone. Advances in optics, miniaturisation, and AI processing are enabling the development of contextual assistants embedded directly into eyewear. For the Android ecosystem, this presents an opportunity to extend AI experiences beyond the handset while reinforcing the broader device portfolio.
Shipments of smart glasses will grow 50% in 2026 to just under 14 million units, according to IDC. While modest compared to smartphone volumes, smart glasses represent a strategic early step toward ambient computing, where the smartphone becomes one node in a larger AI-driven personal network.
Twenty years on
(Image credit: Hayato Huseman / Android Central)
Twenty years after MWC arrived in Barcelona, the smartphone is no longer just a mobile communications device. It is becoming a personal AI endpoint at the center of a broader connected ecosystem. The companies that succeed in this next phase will be those that combine intelligent software, resilient supply chains, and coherent ecosystems — not just compelling spec sheets.
MWC 2026 will not be only about new hardware announcements. It will signal a structural inflexion point.
AI is becoming architectural rather than optional. And yet, despite decades of exponential innovation with faster processors, sharper displays, better cameras, foldable form factors, and intelligence, the industry finds itself constrained by something far more fundamental.
A component long taken for granted is now dictating the pace and economics of innovation. For years, the assumption was clear: component prices, particularly DRAM and NAND, would decline predictably. This assumption no longer holds. Supply chain reallocation toward AI infrastructure has tightened availability and costs.
After decades of engineering breakthroughs, the smartphone industry is being reminded of a simple truth: technological progress is only as strong as the supply chain beneath it.
So my main question heading into MWC 2026 discussions is simple: will competitive advantage come from software intelligence or supply chain resilience?
About IDC
IDC is a leading global provider of technology research, insights, and events. With a network of more than 1,000 analysts across over 100 countries, IDC delivers trusted intelligence on tech markets, trends, and opportunities to help businesses and IT leaders make informed decisions. Learn more today at IDC.com.