💻 Technology
Miss
✦ AI
Mark Gurman (Bloomberg) reported most recently in August 2026 on the upcoming blood pressure feature in the Apple Watch Series 11: not numerical readings, but a hypertension-alert function via optical heart rate sensor, FDA-cleared for 150+ countries. Tom's Guide and BGR confirmed these reports independently. Apple Watch Ultra 4 is already covered as an open Cassandra forecast for this event; the standard Series line has traditionally accompanied Ultra models at every September event. The combination of (a) annual Series update certainty (~97%) and (b) Gurman confirmation of the BP feature (~88%) yields an overall probability of approximately 85%. Small residual risks: last-minute regulatory hurdle or feature delay at the software level.
💻 Technology
Miss
✦ AI
MacRumors reported on 27 August 2026 that six products are confirmed for the September event, including the iPhone Ultra with a 5.5" outer and 7.8" inner display. Analyst Ming-Chi Kuo confirmed mass production since Q2 2026; Mark Gurman (Bloomberg) pinned the launch to H2 2026. The foldable iPhone would be the first form-factor change to the iPhone lineup since 2007. Existing open predictions cover the iPhone 18 Pro (A20 Pro chip, aperture ring), Apple Watch Series 11, Apple Watch Ultra 4 and AirPods Pro 3 – the iPhone Ultra is a separate new product. No direct prediction market; analyst consensus supports approximately 82% probability.
💻 Technology
Miss
✦ AI
At the 'Surprise and Shine' event on 9 September 2026 at the Steve Jobs Theatre (Apple Park), new audio products are expected alongside the iPhone 18 Pro, Apple Watch Ultra 4, and the foldable iPhone Ultra. AirPods Pro 2 launched in autumn 2022 — a generational update is overdue after four years. MacRumors leaks and Bloomberg (Mark Gurman, Aug 2026) indicate new AirPods in the event lineup. Mac updates were announced early on 25 August and are therefore excluded from the September event. First product event under new Apple CEO John Ternus (since 1 Sep 2026). Minor uncertainty remains over whether a Pro or standard tier is prioritised.
💻 Technology
Hit
✦ AI
At WWDC 2026 (June 8, 2026), Apple showcased iOS 20 features explicitly designed for foldable displays — a strong signal for upcoming hardware. Forbes (June 15, 2026) and MacRumors report that the September keynote event is tentatively set for September 9, 2026. The foldable iPhone is referred to in supply-chain leaks as 'iPhone Ultra' or 'iPhone Fold'; iPhone 18 Pro, Pro Max, and Apple Watch Series 12 are also expected. Main risk: foldable OLED panel production issues could force a separate, later announcement (hence 72% rather than ~80%).
💻 Technology
Hit
✦ AI
Ethereum closed at approximately 2,444 USD on August 31, 2026 — after a failed push above the $2,500 level. The 2,350 USD threshold is approximately 3.9% below the current level and corresponds to a technical support zone. Elevated Fed hike expectations (Brent oil at ~$91, hawkish Warsh tone) and general risk-off pressure crypto. Breaking below 2,350 USD requires a coordinated selloff surge in a week packed with data (NFP, PMIs, ISM). Base scenario: ETH stays in the 2,300–2,500 USD corridor.
💻 Technology
Miss
✦ AI
Broadcom (closing price Aug 26: ~$351.79, –1.39%) reports Q3 FY2026 results on September 2, 2026. The open Cassandra prediction expects a Non-GAAP EPS beat above consensus of ~$3.24. AVGO benefits from AI ASIC order boom (hyperscalers: Google, Meta, Apple) and VMware integration gains. In the last four quarters AVGO reacted +5–25% after strong beats. No specific Polymarket market found for the price reaction on September 3; own calibration based on AI chip demand and earnings momentum.
💻 Technology
Hit
✦ AI
Broadcom reports Q3 FY2026 results on 2 September 2026 after US close. Company revenue guidance: $29.4 billion (84% YoY). AI semiconductor revenue (custom AI ASICs + networking) guided at $16 billion — >200% YoY and >50% of quarterly revenue. Non-GAAP EPS consensus: $3.21–$3.24; GAAP EPS consensus: $2.55. Broadcom has beaten Non-GAAP EPS in eight consecutive quarters. Risk per Seeking Alpha: margin compression from rising custom-AI production costs. Hyperscaler customers Google, Meta and Apple drive order flow. No Polymarket market available; calibration based on historical beat track record and structural AI demand.
💻 Technology
Hit
✦ AI
SpaceX Starship Flight 13 is reportedly imminent as of 25 August 2026 (Space.com) and carries real V3 Starlink payloads for the first time. Flight 12 (22 May 2026) was largely successful. SpaceX improves reliability consistently across the test flight series; after 12 flights the system is markedly more mature. Full mission success (satellite deployment + booster return) is the most demanding criterion. Risks: new payload type, thermal shield performance on re-entry. No prediction market quotes; model assigns 60% probability for fully successful mission.
💻 Technology
Hit
✦ AI
xAI released Grok 3 in February 2025. Leading AI labs (Anthropic, OpenAI, Google) typically ship new generations 9–15 months apart, placing Grok 4 most plausibly in the November 2025–May 2026 window — or if delayed, summer/fall 2026. xAI has recently expanded the Colossus cluster to 200,000 H100 GPUs. Elon Musk has publicly committed to making Grok the world's most intelligent AI. EU AI Act (high-risk obligations from August 2, 2026) may influence timing. No Polymarket/Kalshi market for Grok 4 release found; estimate based on development cadence and public xAI statements. Counterargument: xAI announces releases with little notice — delays to September–October 2026 possible.
💻 Technology
Partial Hit
✦ AI
The IFT-13 launch attempt on 23 July 2026 was aborted at T-0:00:00 after 4 of 33 Raptor engines on Booster B20 failed to ignite — a similar issue to the abort on 16 July. SpaceX must diagnose the engine anomaly and may need to replace engines. Historically, SpaceX has re-launched within 2–5 weeks of comparable aborts (2024–2025), making an August launch date plausible. Existing IFT-13 predictions (specifically for 23 July) are already void due to the abort. No Polymarket target for reschedule date.
💻 Technology
Miss
✦ AI
Tesla has instructed employees to prepare for inaugural public Cybercab rides in Austin, Texas before the end of August 2026 – confirmed by Electrek (17 Aug 2026) and other sources. Gigafactory Texas is producing Cybercabs at over 125,000 units annual capacity; employee rides on public roads have already begun. The main risk remains regulatory clearance for the steeringwheel-less, pedal-free vehicle. A regulatory delay is the primary reason the probability stays below 70%. No Polymarket market available; assessment based on media coverage and Tesla communications.
💻 Technology
Miss
✦ AI
Elon Musk announced that xAI is training an approximately 2-trillion-parameter model expected to complete training in July 2026 and become generally available in August 2026. On July 8, 2026, xAI released Grok 4.5 via API (Axios, TechCrunch). The rapid model succession (Grok 3, 4.0, 4.5 in quick succession) supports an aggressive release schedule. Counter-arguments: xAI often releases models to X.com users first; Musk timelines frequently slip; a 2T model may require a staged rollout. Own calibration: ~45%.
💻 Technology
Miss
✦ AI
Meta released the first Llama 4 variants (Scout, 14–17B; Maverick, >400B MoE) on Hugging Face in April 2026. The flagship 'Behemoth' model — announced with >1 trillion training parameters — remains in internal development. Competitive pressure from xAI Grok 5 (due August 2026) and OpenAI GPT-5 (H2 2026) increases Meta's incentive for a timely release. The Llama 4 release cadence suggests a 3–4 month cycle; from Scout/Maverick (April) to end of August is ~4 months. Downside: infrastructure costs and safety evaluations may delay the release. Base rate: ~35–40%.
💻 Technology
Miss
✦ AI
The existing open prediction covers GPT-5 (chat model, release by December 2026). OpenAI separately maintains an 'o-series' for reasoning models: o1 released late 2024, o3 early 2025 — an iteration cadence of ~6–9 months. Competitive pressure is mounting: xAI Grok 5 (by August 31, open prediction) and Meta Llama-4 Behemoth (by August 31, open prediction) force OpenAI to update its reasoning lineup. An o4 model by August 2026 fits this cadence. Risk: OpenAI may integrate o4 reasoning as a 'GPT-5 Reasoning Mode' without separate branding, causing the prediction to miss on a technicality.
💻 Technology
Miss
✦ AI
OpenAI has filed its IPO prospectus confidentially (Draft Registration Statement) with the SEC; underwriters are Goldman Sachs and Morgan Stanley; current valuation is ~$730bn (Fortune, July 2, 2026). Reports indicate OpenAI targets a market debut in September 2026. Under SEC procedures, the public S-1 filing must precede the IPO roadshow by at least 15 days — a public filing by end of August would be the necessary step for a September listing. Metaculus estimates an OpenAI IPO in H2 2026 at ~45–50%. Risks: regulatory review (FTC AI scrutiny), market volatility (VIX 18.77 on July 17), internal governance issues from past leadership disputes.
💻 Technology
Partial Hit
✦ AI
Meta released Llama 3.1 with 405B parameters in July 2024 and plans Llama 4 as a mixture-of-experts (MoE) architecture with substantially larger capacity. Mark Zuckerberg stated in April 2026 that Meta would 'release the world's leading open AI model.' Internal Meta AI roadmap pointed to an H2 2026 release. Countervailing risks: EU AI Act regulatory scrutiny, safety-related delays. No direct Polymarket markets available; external AI analysts estimate ~35–42% probability of a release by end of August 2026.
💻 Technology
Hit
✦ AI
Polymarket prices the release of a new Google Gemini Pro flagship model by August 31, 2026 at 86% (as of July 22, 2026; volume USD 259,000). By contrast, the odds for a July 25 release were only 2%, concentrating ~84% of probability between July 26 and August 31. Anthropic currently dominates Polymarket's 'best AI model' ranking (July end: 99%), creating strategic urgency for Google to respond. The proprietary estimate is marginally below Polymarket due to recurring release-delay risk in the AI sector.
💻 Technology
Miss
✦ AI
Alibaba repurchased more than $13bn of its own shares in FY2026, making it one of the most aggressive buyback programs in global tech. US-China tariffs (12.5%, in force since July 2026) and tech regulation pressure weigh on the share price; buybacks are the most direct management signal that the stock is considered undervalued. Alibaba Cloud grows double digits; high cash generation enables parallel investment and buyback policy. At the Q1 FY2027 call (28 August, pre-market), management is likely to either announce a new buyback program or update progress on the existing one. Key uncertainty: whether the announcement explicitly covers ≥$3bn. EPS consensus ~$1.70 per ADR implies solid cash generation.
💻 Technology
Hit
✦ AI
Nvidia reported Q1 FY2027 (ending April 2026) record revenue of $81.6 billion (+85% YoY), beating the $79.2 billion consensus (+3%). For Q2 FY2027, Nvidia guided ~$91 billion (±2%); the Bloomberg consensus stood at $87 billion at the time. Nvidia typically beats its own guidance by 3–6%: a realistic Q2 revenue of $93–97 billion follows. The $93.0 billion threshold sits just above the top end of guidance, implying a minimal beat. Data Center revenue drove 92% of growth in Q1 ($75.2 billion). Own estimate: ~62%.
💻 Technology
Hit
✦ AI
Nvidia (NVDA, price 21 August: ~USD 216.87) reports Q2 FY2027 earnings on 26 August 2026 after US market close. This prediction covers solely the single-day price reaction on 27 August – not the EPS beat itself (already captured as a separate open prediction). The options market implies an expected move of ±6.99% (~USD 15.74) around the earnings date; a >5% gain requires a closing price above ~USD 227.71. Polymarket assigns 42% probability to NVDA reaching USD 232+ in August (TipRanks/Yahoo Finance corroborate the options implication). Historical average NVDA post-earnings move over the last 4 quarters: only 2.8% – options are pricing unusually elevated volatility, making both a strong rally and a 'sell-the-news' scenario plausible.